Clinical outcomes of totally thoracoscopic mitral valve reoperation with moderate hypothermia-induced ventricular fibrillation arrest

To the Editor: Moderate hypothermia-induced ventricular fibrillation arrest has gained increasing attention in minimally invasive surgery. It has been applied not only in minimal-incision and thoracoscopic procedures, but also, more recently, in robotic surgery.^[1–4] As the number and variety of totally thoracoscopic procedures we performed increase, we have successfully combined this approach with moderate hypothermia-induced ventricular fibrillatory arrest for cardiac reoperations. Although studies have explored the application of moderate hypothermia-induced ventricular fibrillation arrest in minimally invasive surgery,^[5–7] there is a notable lack of research in comparing the totally thoracoscopic approach with the conventional median sternotomy approach. This study retrospectively examined and analyzed the perioperative outcomes of these two approaches using propensity score matching (PSM).

We retrospectively enrolled patients who underwent mitral valve reoperation at the Department of Cardiovascular Surgery, the First Medical Center of PLA General Hospital between January 2008 and December 2023. All surgeries in the study were performed by the same surgeon. The study adhered to the Declaration of Helsinki (revised in 2013) and was approved by Ethics Committee of Chinese PLA General Hospital (No. S2022-343-01), and informed consent was waived with the approval of the Ethics Committee.

The patients were categorized into two groups based on the surgical approach: (1) the totally thoracoscopic approach with moderate hypothermia-induced ventricular fibrillation arrest (TT group), and (2) the conventional median sternotomy approach (MS group). The inclusion criteria included a history of at least one prior cardiac surgery and presentation with new-onset mitral valve disease requiring surgical treatment, such as bioprosthetic valve failure, mechanical valve dysfunction, infective endocarditis, mitral valve perivalvular leak, recurrent dysfunction after mitral valvuloplasty, or new-onset severe mitral valve regurgitation or stenosis. The exclusion criteria included a history of mitral valve balloon dilatation, the need for concurrent aortic surgery, aortic valve surgery, or coronary artery bypass grafting, and the presence of moderate or severe aortic regurgitation. Preoperative ultrasound of the femoral, iliac, and jugular arteries was routinely performed to assess cannula diameter and exclude vascular diseases that could complicate cardiopulmonary bypass (CPB) establishment. Postoperative echocardiography was conducted approximately one week postoperatively to evaluate cardiac function.

Intraoperative CPB was established via the femoral artery and right internal jugular vein. A 2.0- to 3.0-cm incision was made in the groin for femoral artery and vein access, where two purse-string 5-0 nonabsorbable sutures were applied to the femoral artery and one to the femoral vein. Direct cannulation using the Seldinger technique employed a 15 F to 18 F (1 F = 0.33 mm) tube for arterial cannulation and a 21 F or 23 F tube for venous cannulation as the inferior vena cava line, whereas the right internal jugular vein was cannulated with a 15 F tube for the superior vena cava line. Negative-pressure suction was applied to the venous cannula, and its end was positioned at the fourth intercostal space and adjusted as necessary. A 2.5- to 3.0-cm skin incision was made in the lateral right nipple area in men and the right inframammary fold in women, with entry into the chest through the lateral fourth intercostal space in the right midclavicular line, serving as the primary operative site. The thoracoscope was placed through the fourth intercostal space in the anterior axillary line, and a carbon dioxide tube was introduced into the chest cavity via a skin protector. Surgical access was adjusted if extensive severe pulmonary pleural adhesions were present, although mild adhesions typically did not require changes. A left heart drain was inserted via puncture in the fifth intercostal space in the midaxillary line, while a left atrial pull-hook was positioned through punctures in the third and fourth intercostal spaces on the right side of the sternum. During CPB, the temperature was gradually reduced, and the left atrium was incised parallel to the interatrial sulcus upon the onset of ventricular fibrillation to expose the mitral valve for surgery. After the procedure, the temperature was gradually raised, and sinus rhythm was restored using extracorporeal electrode plate defibrillation. Following CPB cessation, the CPB tube remained in place temporarily, and transesophageal ultrasound was used to evaluate the surgical outcome.

The baseline characteristics and perioperative outcomes were collected from the electronic medical system. Baseline characteristics included age, sex, body mass index, hypertension, diabetes mellitus, stroke history, peripheral arterial calcification, coronary artery disease, atrial fibrillation, New York Heart Association classification, chronic lung disease, left atrial dimension (LAd), left ventricular end-diastolic dimension (LVDd), left ventricular ejection fraction (LVEF), blood creatinine level, and platelet count. Perioperative outcomes included the CPB time, ventricular fibrillation or aortic cross-clamping time, minimum rectal temperature during surgery; transfusion rate, postoperative LAd, LVDd, and LVEF, postoperative troponin-T and creatine kinase isoenzyme-MB (CK-MB) levels, ventilator time, intensive care unit (ICU) stay, drainage volume, postoperative hospital stay, intra-aortic balloon pumping and extracorporeal membrane oxygenation implantation, secondary thoracotomy, continuous renal replacement therapy, stroke or transient ischemic attack (TIA), and mortality. Serum troponin-T and CK-MB levels were measured after postoperative transfer to the ICU.

Statistical analysis was conducted using SPSS 22.0 (IBM Corp., Armonk, NY, USA) and R 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables are expressed as mean ± standard deviation or median (Q₁–Q₃), whereas categorical data are presented as frequency and percentage. Comparisons between the two groups were performed using Student’s t-test for normally distributed continuous variables, the Mann–Whitney U test for skewed continuous variables, and the χ² test or Fisher’s exact test for categorical variables. Multiple linear regression was applied to clarify the relationship between arrest time and serum myocardial enzyme levels. A P-value of <0.05 was considered statistically significant. To minimize selection bias, the PSM method was used. Propensity scores were calculated using all baseline characteristics, including age, sex, body mass index, hypertension, diabetes mellitus, stroke history, peripheral arterial calcification, coronary artery disease, atrial fibrillation, New York Heart Association classification, chronic lung disease, LAd, LVDd, LVEF, blood creatinine level, and platelet count. Patients in the TT group were matched to those in the MS group using a caliper width of 0.25 standard deviations of the propensity score.

After the application of the inclusion and exclusion criteria, 181 patients were enrolled (64 in the TT group and 117 in the MS group). The prior surgical procedures for patients in the TT and MS groups were valve surgery (65.6% [42/64] vs. 77.8% [91/117], P = 0.077), coronary artery bypass grafting (7.8% [5/64] vs. 2.6% [3/117], P = 0.101), and congenital heart disease surgery or other procedures (26.6% [17/64] vs. 19.7% [23/117], P = 0.284). The indications for reoperation in the TT and MS groups were new-onset mitral valve disease (42.2% [27/64] vs. 43.6% [51/117], P = 0.855), prosthetic valve dysfunction (34.4% [22/64] vs. 33.3% [39/117], P = 0.887), perivalvular leakage of prosthetic valves (15.6% [10/64] vs. 18.0% [21/117], P = 0.692), and infective endocarditis (7.8% [5/64] vs. 5.1% [6/117], P = 0.470).

After PSM, the sample size was balanced with 51 patients in each group, and the PSM procedure achieved a good covariate balance between the two groups. The distribution of propensity scores for both groups is presented in Supplementary Figure 1, http://links.lww.com/CM9/C462. Before PSM, significant differences were observed in age (58.2 ± 15.8 years vs. 52.7 ± 16.3 years, P = 0.028), peripheral arterial calcification (48.4% [31/64] vs. 23.9% [28/117], P <0.001), and platelet count ([195.7 ± 58.2] × 10⁹/L vs. [169.6 ± 65.3] × 10⁹/L, P = 0.006) between the two groups. However, after PSM, no statistically significant differences were found in any of the variables between the groups, as detailed in Supplementary Table 1, http://links.lww.com/CM9/C462.

After PSM, the TT group had a longer CPB time (190.2 ± 54.3 min vs. 135.1 ± 40.8 min, P <0.001), whereas the intraoperative ventricular fibrillation time in the TT group was 97.6 ± 41.8 min, and the aortic cross-clamping time in the MS group was 84.2 ± 31.8 min. The TT group had a lower minimum rectal temperature during surgery (27.3 ± 1.4°C vs. 31.3 ± 1.5°C, P <0.001), a lower transfusion rate (51.0% vs. 100%, P <0.001), higher postoperative LVEF ([61.1 ± 6.7]% vs. [57.6 ± 8.0]%, P = 0.019), shorter ventilator time (13.0 [10.5–18.3] h vs. 17.5 [16.8–25.5] h, P <0.001), lower drainage volume (180.0 [120.0–585.0] mL vs. 900.0 [700.0–1455.0] mL, P <0.001), and shorter postoperative hospital stay (8.0 [7.0–10.0] days vs. 10.0 [8.0–13.0] days, P = 0.005). There were no statistically significant differences between the two groups in postoperative LAd, LVDd, postoperative troponin-T level, CK-MB level, ICU stay, intra-aortic balloon pumping and extracorporeal membrane oxygenation implantation, secondary thoracotomy, continuous renal replacement therapy, stroke/TIA, or mortality, as shown in Supplementary Table 2, http://links.lww.com/CM9/C462.

Compared with first-time surgery, mitral valve reoperation is more challenging because of its higher complication rate and mortality.^[8–10] The conventional median sternotomy approach poses risks of damaging critical structures such as the bridge vessels behind the sternum, ascending aorta, and right ventricle. Complications such as extensive bleeding from adhesive tissue separation, sternal wound infection, and sternal osteomyelitis associated with sternotomy are significant concerns. Surgeons often need to dissect the ascending aorta for occlusion, the right side of the heart to establish CPB, and the left side to enhance visualization of the mitral valve. This extensive dissection increases the risk of injury and hemorrhage in vital cardiac structures. The totally thoracoscopic approach with moderate hypothermia-induced ventricular fibrillation arrest offers advantages by bypassing the severe adhesions encountered in a previous median thoracotomy. Approaching from the right side of the chest minimizes intraoperative trauma and reduces procedural difficulty.

This study has some limitations. Our study was based on a single-center experience and involved a relatively small sample size. Although the PSM technique effectively balanced the groups to simulate randomization, unmeasured factors could still confound the results. Moving forward, we plan to adopt a medium- to long-term follow-up for the patients included in this study.

In conclusion, this study demonstrates that totally thoracoscopic mitral valve reoperation with moderate hypothermia-induced ventricular fibrillation arrest does not increase the incidence of stroke/TIA or mortality compared with conventional surgery and offers significant advantages in terms of reduced trauma and faster recovery.

Conflicts of interest

None.

Supplementary Material

cm9-138-1761-s001.docx ^{(131.2KB, docx)}