Minimally Invasive Cardiac Surgery in China: Multi-Center Experience.

BACKGROUND To the best of our knowledge, there is no nationwide data available on the development of minimally invasive cardiac surgery (MICS) in China. The purpose of this study was to report the results of MICS in 6 experienced centers in China. MATERIAL AND METHODS From September 2014 to July 2016, 1241 patients with cardiac conditions who underwent MICS procedures were randomly enrolled in 6 centers in China, and those patients were randomly selected for inclusion in this study. The MICS procedures were defined as any cardiac surgery performed through a less invasive incision, rather than a complete median sternotomy, mainly including mini-incision surgery (400, 32.2%), video-assisted approach (265, 21.3%), completely thoracoscopic approach without robotic assistance (504, 40.6%), and robotic procedure (55, 4.4%). RESULTS The 5 most common in-hospital complications were respiratory failure (28, 2.3%), reoperation for all reasons (19, 1.5%), renal failure (11, 0.9%), heart failure (9, 0.7%), and stroke (6, 0.5%). The multivariate logistic regression analysis results showed that cardiopulmonary bypass (CPB) time (P=0.033), aortic cross-clamp time (P=0.003), cannulation approach (P=0.010), and left ventricular ejection fraction (LVEF) (P=0.003) at baseline were all significant risk factors of any in-hospital complication of MICS procedures. CONCLUSIONS From our experience, minimally invasive cardiac approaches are safe and reproducible, with acceptable CPB and aortic cross-clamp time duration and low mortality.

Background

Since the 1990s, minimally invasive cardiac surgery (MICS) has had substantial progress and innovation [1,2]. A variety of MICS procedures have been reported in the literature, including minithoracotomy [3-6], hemi-sternotomy [7], video-assisted approach [8], completely thoracoscopic approach [9], robotic procedure [10], and catheter-based hybrid approach [11]. The results suggest that MICS is associated with the potential benefit of improved cosmesis, less pain, shorter recovery duration, and faster return to normal activities, without compromising the safety of the procedure [12,13]. Furthermore, MICS has been used in the treatment of various cardiac lesions, and has become the preferred treatment standard in some heart centers [14-17]. MICS is becoming important in the field of cardiac surgery, and be applied globally. With the dramatic development of cardiac surgery in China, numerous surgeons are committed to MICS techniques [9,18-22].

To the best of our knowledge, there is no nationwide data available on the development of MICS in China. The aim of the current study was to report the results of minimally invasive cardiac surgery (MICS) based on the experiences of 6 centers in China.

Material and Methods

Patients

Between September 2014 and July 2016, MICS procedures were performed in 1241 patients with coronary artery disease, valve disease, congenital heart disease, and atrial fibrillation in 6 centers in China. Those patients were randomly enrolled in this study. The MICS procedures were defined as any cardiac surgery performed through a less invasive incision, rather than a complete median sternotomy, including minithoracotomy, hemisternotomy, video-assisted approach, completely thoracoscopic surgery without robotic assistance, and robotic procedure.

Inclusion criteria were as follows: (1) 4 years old or more, weighing 14 kg or more; (2) no severe peripheral vascular disease; (3) no preoperative history on the right hemithorax; (4) no history of pleural infection and severe adhesion of pleurae; (5) no severe deformity of trachea and bronchus; (6) no mental retardation, and (7) no severe coagulopathy.

The Ethics Review Board of Beijing Anzhen Hospital, Capital Medical University approved this research project. The methods were carried out in accordance with the relevant guidelines and written informed consent was provided by all participants.

Statistical analysis

Quantitative variables are expressed as means± standard deviations, and categorical variables as frequencies and percentages. Multivariate logistic regression (backward stepwise, remove P>0.15) was used to assess the impact of independent variables on complications during hospitalization, including respiratory failure, reoperation for all reasons, renal failure, heart failure, stroke, and acute aortic dissection after MICS. Adjusted variables included demographics (gender, age, BMI), comorbidities (left ventricular ejection faction, blood creatinine, hypertension, atrial fibrillation, current smoke, hypercholesterolemia, previous MI, cerebrovascular disease, COPD, diabetes, etc) were known to be risk factors for surgical procedure. Age, BMI, left ventricular ejection faction, atrial fibrillation, and blood creatinine were included as continuous variables. Gender, hypertension, current smoke, hypercholesterolemia, previous MI, cerebrovascular disease, COPD, and diabetes were included as categorical variables. In this regression analysis, we did not analyze each of these complications because the incidence of individual complications was low and the sample size was limited and could not be effectively analyzed by logistic regression. Therefore, we added all the complications and integrated them into a dependent variable Y, for analysis. Predictors of choice consider the following 2 aspects: (1) statistical considerations, first by single-factor analysis if the p value was less than 0.05 for the variables included in the model; and (2) clinical considerations, even if the single-factor analysis P value was greater than 0.05, but the clinician thought those were important variables, we also included them in the model. Demographic data, comorbidities, and other factors mentioned in the article were evaluated. Model discrimination was assessed by calculating the area under the receiver operating curve (AUC). All P values of less than 0.05 were considered to be statistically significant. All reported P values are 2-sided. All data were analyzed with the SPSS software package for Windows (version 15.0; SPSS, Chicago, IL).

Results

A total of 1241 patients with cardiac conditions received a variety of MICSs, including mini-incision surgery, video-assisted approach, completely thoracoscopic approach without robotic assistance, and robotic procedure. Baseline characteristics and demographics of patients are listed in Table 1.

Table 1

Baseline characteristics and demographics of patients undergoing minimally invasive cardiac surgery*.

Variable	Minimally invasive cardiac surgery (n=1241)
Age, years	47.9±16.8
≥70	78 (6.3%)
≥ 80	3 (0.2%)
Gender
Male	650 (52.4%)
BMI (Kg/m2)	22.8±12.0
NYHA class
I/II	997 (80.3%)
III/IV	244 (19.7%)
LVEF (%)	63.8±7.6
Creatinine	65.7±22.4
CHF	13 (1.0%)
Current smoke	106 (8.5%)
Hypertension	130 (10.5%)
Hypercholesterolemia	66 (5.3%)
AF	194 (15.6%)
Previous MI	41 (3.3%)
Cerebrovascular Disease	14 (1.1%)
COPD	13 (1.0%)
Diabetes Mellitus	54 (4.4%)
Previous CVTS	10 (0.8%)

Continuous data are reported as mean (SD); categorical data are presented as number (%).

BMI – body mass index; NYHA – New York Heart Association; LVEF – left ventricular ejection fraction; CHF – chronic heart failure; AF – atrial fibrillation; MI – myocardial infarction; COPD – chronic obstructive pulmonary disease; CVTS – cardiovascular and thoracic surgery.

The mean body mass index (BMI) was 22.8 Kg/m2 (patients with BMI <25 Kg/m2 accounted for 76.1%, BMI between 25 and 35 Kg/m2 were 23.4%, and another 0.5% BMI >35 Kg/m2). Notably, according to the New York Heart Association (NYHA) grade, 80.3% patients belonged to class I or II. Patients had a variety of preoperative concomitant disorders, and the most common were atrial fibrillation in 194 (15.6%) patients and hypertension in 130 (10.5%) patients.

The most common surgical method was completely thoracoscopic approach without robotic assistance (504, 40.6%), followed by mini-incision surgery (400, 32.2%), video-assisted methods (265, 21.3%), and robotic procedure (55, 4.4%). In addition, 17 patients (1.5%) were converted to a full medial sternotomy (Table 2).

Table 2

In-hospital outcomes of all types of minimally invasive cardiac surgeries*.

Variables	Mini-incision (n=400)	Video-assisted (n=265)	Totally thoracoscopic (n=504)	Robotic (n=55)
CPB time (min)	113.1±48.2	116.1±36.3	96.7±65.9	83.8±29.9
Cross-clamp time (min)	74.8±34.6	74.6±27.0	58.9±30.0	54.0±21.2
LOS in ICU (hours)	99.1±36.3	61.3±7.7	33.5±15.6	57.3±10.5
LOS in hospital (days)	25.0±5.5	16.2±1.1	15.4±6.2	15.9±2.1
Hospital stay on POD (days)	13.7±3.6	10.5±2.8	9.0±3.3	11.0±3.5
Ventilatory support (hours)	31.1±23.5	17.8±8.8	13.0±7.5	18.9±4.4
Intraoperative PRBC transfusion	316 (79.2%)	128 (78.0%)	119 (24.3%)	47 (87.0%)
Intraoperative plasma transfusion	326 (81.7%)	131 (79.9%)	110 (22.5%)	38 (70.4%)
Postoperative PRBC transfusion	247 (61.9%)	80 (48.8%)	108 (22.1%)	18 (33.3%)
Postoperative plasma transfusion	334 (83.7%)	113 (68.9%)	136 (27.8%)	32 (59.3%)
Drainage on POD (mL)	1048.5±747.2	826.3±526.8	602.2±460.5	560.0±432.5

Continuous data are reported as mean (SD); categorical data are presented as number (%).

CPB – cardioplumonary bypass; LOS – length of stay; ICU – intensive care unit; BPU – blood products usage; POD – postoperative day; SD – standard deviation.

The 5 most common procedures performed in the enrolled patients were mitral valve repair or replacement (MVR/r) (363, 29.3%), atrial septal defect (ASD) (359, 28.9%), ventricular septal defect (VSD) (141, 11.4%), atrial fibrillation (AF) (121, 9.8%), and coronary artery bypass grafting (CABG) (82, 6.6%) (Table 3). In addition, 175 (14.1%) patients underwent other basic procedures. In-hospital death occurred in 7 patients (0.6%). The 5 most common in-hospital complications were respiratory failure (28, 2.3%), reoperation for all reasons (19, 1.5%), renal failure (11, 0.9%), heart failure (9, 0.7%), and stroke (6, 0.5%). No acute aortic dissections were found in the series.

Table 3

In-hospital outcomes of some common cardiac surgeries*.

Variables	Isolated MVR/r (n=363)	Isolated ASD (n=359)	Isolated VSD (n=141)	Isolated CABG (n=82)
CPB time (min)	107.±46.9	101.±51.4	104.3±48.3	113.8±58.2
Cross-clamp time (min)	68.4±31.4	66.2±37.2	69.0±32.9	70.9±33.6
LOS in ICU (hours)	69.3±58.9	71.1±65.0	64.5±36.1	62.0±40.8
LOS in hospital (days)	31.1±10.7	15.7±1.8	22.3±2.8	40.2±9.3
Hospital stay on POD (days)	20.9±7.1	25.1±7.5	33.9±10.2	27.3±5.6
Ventilatory support (hours)	25.9±37.4	25.3±36.3	23.2±20.6	13.0±7.5
Intraoperative PRBC transfusion	235 (75.1%)	242 (70.8%)	85 (75.9%)	49 (76.6%)
Intraoperative plasma transfusion	217 (69.3%)	220 (64.3%)	88 (78.6%)	49 (76.6%)
Postoperative PRBC transfusion	141 (45.0%)	147 (43.0%)	51 (45.5%)	35 (54.7%)
Postoperative plasma transfusion	183 (58.5%)	214 (62.6%)	66 (58.9%)	35 (54.7%)
Drainage on POD (mL)	817.9±622.2	801.9±595.3	826.3±526.8	904.2±686.7

Continuous data are reported as mean (SD); categorical data are presented as number (%).

MVR/r – mitral valve replacement/repair; ASD – atrial septal defect; VSD – ventricular septal defect; CABG – coronary artery bypass grafting; CPB – cardioplumonary bypass; LOS – length of stay; ICU – intensive care unit; BPU – blood products usage; POD – postoperative day; SD – standard deviation.

Of the 45 institutions that can perform cardiac surgery in China, 23 (51.1%) institutions have the ability to perform the mini-incision cardiac surgery, 15 (33.3%) institutions can perform the thoracoscopic surgery, and 7 (15.6%) institutions can perform robotic surgery.

The multivariate logistic regression analysis results showed that cardiopulmonary bypass (CPB) time (P=0.033), aortic cross-clamp time (P=0.003), cannulation approach (P=0.010), and left ventricular ejection fraction (LVEF) (P=0.003) at baseline were all significant risk factors (Table 4) of all in-hospital complications of MICS procedures. The calculated area under curve (AUC) for the model was 0.775.

Table 4

Multivariate logistic regression analysis of risk factors for in-hospital complications*.

Risk factor	P	Odds ratio	95% CI
CPB time (min)	0.033	1.004	1.001–1.007
Aortic cross-clamp time (min)	0.003	1.012	1.004–1.021
LVEF (%)	0.003	1.072	1.025–1.121
Cannulation approach	0.010	2.208	1.205–4.046

Estimates were adjusted for variables, including demographics (gender, age, BMI), and comorbidities (left ventricular ejection faction, blood creatinine, hypertension, atrial fibrillation, current smoke, hypercholesterolemia, previous MI, cerebrovascular disease, COPD, diabetes) shown in the table.

CI – confidence interval; CPB – cardiopulmonary bypass; LVEF – left ventricular ejection fraction.

Discussion

With the substantial improvement in techniques and instruments, MICS has evolved from mini-incision to thoracoscopy and robotics. It is well understood that MICS does not just refer to single methods, referring instead to a collection of new techniques and instruments [23]. Moreover, with the rapid popularization of some specialized centers, MICS has now become the standard method of cardiac surgery [14,16]. Previous studies have reported that MICS is used in almost all types of cardiac surgery, including mitral valve surgery [23], ASD repair [20,24], VSD repair [25], AF ablation [26], CABG [22,27], and myxoma dissection [28]. Notably, results of randomized clinical trials (RCTs) and meta-analyzes [23] have clearly proven that MICS is a safe, efficacious, and repeatable alternative to traditional surgical procedures. In addition, some retrospective studies reported that, despite the lack of RCTs in meta-analysis, other minimally invasive procedures were still safe and reliable.

In the last several decades, cardiac surgery in China has experienced tremendous development. Some reports from China reflected the domestic situation of MICS [10,21,22,29,30] in China to some extent. However, due to the sparsely populated, single-agency limitations, these reports cannot accurately reflect the status of MICS in China. The present study included 1241 patients who underwent MICS in 6 experienced centers in China, in order to assess the current status of MICS in China. With regard to baseline characteristic and demographics of the patient cohorts, obvious selection bias existed. In this study, only 6.3% of patients were at least 70 years of age, and 0.2% of patients were at least 80 years of age. In addition, nearly 76.1% of patients had a BMI <25 Kg/m2 and only 0.5% of patients had a BMI >35 Kg/m2. This suggests that few MICS were performed in the elderly and the obese populations in China. Notably, according to the New York Heart Association (NYHA) grade, 80.3% patients belong to class I or II; accordingly, the selected patients who underwent MICS in this study had good preoperative clinical conditions.

With regard to the in-hospital outcomes in this study, MICS was performed successfully, with only 1.4% needing conversion to median sternotomy and an in-hospital mortality of less than 0.6%. The conversion rate was consistent with the previous result reported by Iribarne et al. [31], but the in-hospital mortality was lower than Iribarne’s result, perhaps due to patient selection bias. Respiratory failure (2.3%) was the first postoperative complication, which is consistent with previous reports. There was low occurrence of stroke (0.5%) and no acute aortic dissection, which demonstrates the superiority of thoracic aortic clamp method to the endo-aortic occlusion. The mean CPB time and aortic cross-clamp time, sorted either by surgical approach or by cardiac lesions, indicated that MICS can be achieved within an acceptable operating time.

Femoral artery and central aortic catheterization were the most common approaches of cannulation for establishment of CPB in our enrolled institutions. In multivariate logistic regression analysis, we found that increased risk of major in-hospital complications was associated with use of femoral artery cannulation versus central aortic cannulation. In addition, other risk factors associated with major in-hospital complications after MICS (e.g., LVEF, CPB time, and aortic cross-clamp time) were consistent with results previously reported for patients undergoing cardiac surgery through totally median sternotomy [4-6,32].

Limitations

First, most enrolled patients in this study were in good preoperative medical condition. Thus, patient selection bias might exit. In a future study, we will select patients who underwent conventional surgery with totally median sternotomy at the same time, and closely match the characteristics of the MICS series to minimize selection bias.

Second, this study only reflected the in-hospital situation of MICS, and did not assess the intermediate and long-term outcomes. As a result, no further follow-up has yet been made, and the intermediate and long-term results of MICS need further evaluation.

Third, MICS is composed of a variety of surgical approaches used to treat a variety of cardiac lesions. The present study only summarized the current situation of MICS in China. In a subsequent in-depth study, we will investigate a single-mode MICS approach or a single type of cardiac lesion.

Conclusions

MICS has undergone substantial development in China. This early multi-institution study shows that minimally invasive approaches are safe and reproducible, with acceptable CPB and aortic cross-clamp duration and low mortality.