Does concomitant tricuspid valve surgery increase the risks of minimally invasive mitral valve surgery? A multicentre comparison based on data from The Netherlands Heart Registration

Abstract

Objectives

Mitral valve (MV) disease is often accompanied by tricuspid valve (TV) disease. The indication for concomitant TV surgery during primary MV surgery is expected to increase, especially through a minimally invasive surgical (MIS) approach. The aim of the current study is to investigate the safety of the addition of TV surgery to MV surgery in MIMVS in a nationwide registry.

Methods

Patients undergoing atrioventricular valve surgery through sternotomy or MIS between 2013 and 2018 were included. Patients undergoing MV surgery only through sternotomy or MIS were used as comparison. Primary outcomes were short‐term morbidity and mortality and long‐term survival. Propensity score matching was used to correct for potential confounders.

Results

The whole cohort consisted of 2698 patients. A total of 558 patients had atrioventricular double valve surgery through sternotomy and 86 through MIS. As a comparison, 1365 patients underwent MV surgery through sternotomy and 689 patients through MIS. No differences in 30‐ and 120‐day mortality were observed between the groups, both unmatched and matched. 5‐year survival did not differ for double atrioventricular valve surgery through either sternotomy or MIS in the matched population (90.1% vs. 95.3%, Log‐Rank p = .12). A higher incidence of re‐exploration for bleeding (n = 12 [15.2%] vs. n = 3 [3.8%], p = .02) and new onset arrhythmia (n = 35 [44.3%] vs. n = 13 [16.5%], p < .001) was observed in double valve surgery through MIS. Median length of hospital stay (LOHS) was longer in the minimally invasive double valve group (9 days [6–13]) compared with sternotomy (7 days [6–11]; p = .04).

Conclusion

No differences in short‐term mortality and 5‐year survival were observed when tricuspid valve was added to MV surgery in MIS or sternotomy. The addition of tricuspid valve surgery is associated with higher incidence of re‐exploration for bleeding, new onset arrhythmia. A longer LOHS was observed for MIS compared to sternotomy.

Abbreviations

ASD

atrial septal defect

LOHS

length of hospital stay

MIDVS

minimally invasive double valve surgery

MIMVS

minimally invasive mitral valve surgery

MR

mitral regurgitation

MV

mitral valve

PAH

pulmonary artery hypertension

SDV

sternotomy double valve surgery

SMV

sternotomy mitral valve surgery

1. INTRODUCTION

Mitral valve (MV) regurgitation is the second most common valvular heart disease in the Western world. ¹ Although mitral stenosis (MS) and mitral regurgitation (MR) are distinct mitral pathologies, their effect on left atrial pressure, pulmonary hypertension and subsequent right ventricular (RV) afterload are equitable. ² As such, MV pathology significantly affects the RV and its accompanying structures, among others the tricuspid valve (TV). Secondary basal RV dilatation often results in TV leaflet tethering, TV annular dilatation, leaflet malcoaptation and in many cases, secondary tricuspid regurgitation (TR). ³ Interestingly, in 62.5% of patients diagnosed with significant TR, its actual cause was perceived to be left‐sided valve disease. ⁴ Due to this inseparable link between the atrioventricular valves, it is imperative to adequately evaluate the anatomy and function of the TV, when surgery for MV disease is indicated. As the prevalence of MV disease and MV surgery is increasing, ⁵ this also applies to the presence or prevention of secondary TR. Indeed, current guidelines do not only recommend concomitant TV surgery when secondary severe TR is present, but also in moderate TR accompanied by TV dilatation (>40 mm). ⁶ , ⁷ Moreover, new evidence–not yet incorporated in guidelines‐even suggests to perform concomitant TV repair (TVr) in patients with mild TR and significant TV dilatation, ⁸ as many patients with TV dilatation eventually progress to severe TR and subsequent RV failure, irrespective of durable MV surgery.

Based on these developments, the indication for concomitant TV surgery during primary MV surgery is expected to increase. For isolated MV surgery, minimally invasive mitral valve surgery (MIMVS) has proven to be a safe and effective alternative to sternotomy, in terms of mortality, complication rate and efficacy. ⁹ However, data is scarce on the outcomes of concomitant TV surgery during MIMVS, and surgeons may be apprehended to perform concomitant TV surgery during MIMVS due to the associated increased cardiopulmonary bypass (CPB)‐ and aortic clamping times. ¹⁰ , ¹¹ Therefore, the current study aims to investigate the outcomes of the addition of TV surgery to MV surgery performed through sternotomy and through a minimally invasive surgical (MIS) approach, using a nationwide multi‐centre registry.

2. PATIENT AND METHODS

2.1. Study design

The current retrospective multicentre cohort included patients between January 2013 and December 2018 from 16 cardiothoracic centers in the Netherlands. To evaluate the outcomes of the addition of TV surgery to MV surgery performed through sternotomy and through MIS, two separate questions were formulated:

• − Does the addition of concomitant TV surgery add additional risk of mortality to MIMVS (safety)?
• − Are there differences in short‐ and Midterm outcomes between double atrioventricular valve surgery performed through sternotomy or MIS (safety and efficacy)?

Therefore, the study population was divided in four subgroups, namely

• –(1) MV surgery through sternotomy
• –(2) MV + TV through sternotomy
• –(3) MV surgery through anterolateral mini‐thoracotomy
• –(4) MV + TV surgery through anterolateral mini‐thoracotomy

2.2. Source of study data

The data of the current study are retrieved from the nationwide mandatory prospective database of the Netherlands Heart Registration (Nederlandse Hart Registratie, NHR). Results on primary MV and reoperative MV surgery from this nationwide registry were previously reported by our research group. All data are anonymized for both patients, surgeons and centers. All mandatory variables for the NHR registry were complete. However, two nonmandatory variables had a relatively high proportion of missing data: cardiopulmonary bypass time and aortic cross‐clamp time Given the higher levels of missing data for these variables, which clearly exceed >10% and most likely had a nonrandom distribution, it was inappropriate to apply a multiple imputation method, impairing analysis of these surgical times.

2.3. Inclusion

Adult patients who operated for mitral valve surgery and patients with double atrioventricular valve surgery were included. Mitral valve surgery was defined as either isolated MV repair (MVr) or isolated replacement (MVR) or MVr/MVR with rhythm surgery (pulmonary vein isolation/MAZE procedure) and/or atrial septal closure. Double atrioventricular valve surgery was defined as MV surgery with either tricuspid valve repair (TVr) as tricuspid valve replacement (TVR) with or without rhythm surgery (pulmonary vein isolation/MAZE procedure) and/or atrial septal closure. Patients undergoing other concomitant interventions were excluded. Minimally invasive surgery (MIS) was defined as right‐sided anterolateral thoracotomy with peripheral cannulation, avoiding (re)sternotomy. Robotic mitral valve surgery was excluded.

2.4. Outcomes

Short‐term outcomes, defined as early mortality (30‐day mortality, 120‐day mortality) and postoperative complications and duration of hospital stay, were retrieved. Definitions regarding postoperative complications are presented in Appendix. Midterm outcomes were defined as midterm survival at 5 years. Survival data were derived from the municipal administration records and were completed for all patients. Survival follow‐up was completed through December 1, 2020.

2.5. Statistical analysis

The normality of the continuous variables was tested by visual inspection of the histograms and the Shapiro–Wilk test. Continuous data are presented as mean ± standard deviation or as median with [interquartile range] in the presence of skewness. Categorical data are expressed as frequencies and percentages and were compared using the χ² test and Fisher's Exact test when the minimum count was not met. Continuous variables were compared using the t‐test in case of normality and the Mann–Whitney U‐test in case of skewed data. Kaplan–Meier survival curves were used to demonstrate midterm survival and the Log‐Rank test was applied to assess differences.

Propensity‐score matching analyses were performed to compare minimally invasive double atrioventricular valve surgery to minimally invasive mitral valve surgery and double atrioventricular valve surgery through sternotomy. Propensity scores were estimated using covariates identified at baseline differences. Propensity scores were matched using nearest neighbor matching in a 1:1 ratio, replacement was not allowed, with a 0.01 calliper. Baseline characteristics were presented before and after propensity score matching. Standardized mean difference (SMD) were used to compare the difference in means in units of the pooled standard deviation. A value higher than 0.10 was considered an index of residual imbalance.

All reported p‐values were two‐sided and were considered statistically significant when p < .05. Statistical analyses were performed using SPSS software (V28, IBM) and R Statistics (the R Foundation).

3. RESULTS

3.1. Minimally invasive double valve surgery compared to minimally invasive single valve surgery

Patient referred for MIDVS were older (73 [67–77] vs. 64 [56–72]; p < .001) and less frequently male (n = 34 [39.5%] vs. n = 415 [60.2%]; p < .001) compared with MIMVS (Table 1). A higher incidence of diabetes (n = 9 [10.5] vs. n = 26 [3.8]; p = .01) was observed in the MIDVS group. Overall, the logistic EuroSCORE was significantly higher in the MIDVS group (5.5 [4.2–7.7] vs. 2.9 [1.8–5.5]; p < .001). In the MIMVS group more MVr was performed (n = 511 [74.2] vs. n = 54 [62.8]; p = .03) resulting in a higher repair rate (77.4% vs. 63.7%; p = .01) (Table 2). No differences in 30‐day mortality (n = 0 [0] vs. n = 5 [0.7]; p = .99) and 120‐day mortality (n = 1 [1.2] vs. n = 7 [1.0]; p = .99) were observed (Table S1). A longer median hospital stay (9 days [6–13] vs. 6 days [5–8]; p < .001), a higher incidence of new onset arrythmia (n = 39 [45.3] vs. n = 134 [19.4]; p < .001) and higher need for re‐exploration of bleeding (n = 12 [14.0] vs. n = 30 [4.4]; p = .001) was observed for MIDVS in the unmatched population compared with MIMVS (see Table S1).

Table 1.

Baseline characteristics of all unmatched subgroups

	Sternotomy		Minimally invasive		p‐value
	MV N = 1365	MV + TV N = 558	MV N = 689	MV + TV N = 86	SMV vs. SDV	MIMVS vs. MIDVS	SDV vs. MIDVS
Age, median [IQR]	64 [55–72]	70 [62–75]	64 [56–72]	73 [67–77]	<.001	<.001	<.001
Male sex, n (%)	792 (58.0)	281 (50.4)	415 (60.2)	34 (39.5)	.002	<.001	.06
Diabetes, n (%)	122 (8.9)	46 (8.2)	26 (3.8)	9 (10.5)	.63	.01	.49
Chronic Lung Disease, n (%)	134 (9.8)	74 (13.3)	30 (4.4)	8 (9.3)	.03	.06	.31
Extracardiac arteriopathy, n (%)	46 (3.4)	26 (4.7)	12 (1.7)	1 (1.2)	.18	.99	.24
Recent myocardial infarction, n (%)	7 (0.5)	9 (1.6)	3 (0.4)	0 (0)	.02	.99	.62
Active endocarditis, n (%)	34 (2.5)	0 (0)	2 (0.3)	0 (0)	<.001	.99	–
Serum creatinine (>200 μm/L)	15 (1.1)	11 (2.0)	2 (0.3)	0 (0)	.13	.99	0.38
LVEF, median [IQR]	55 [50–60]	55 [40–55]	55 [55]	55 [50–55]	<.001	.65	.004
Good, n (%)	978 (71.6)	298 (53.4)	586 (85.1)	64 (74.4)	<.001	.01	<.001
Moderate, n (%)	375 (27.5)	239 (42.8)	91 (13.2)	23 (24.4)	<.001	.01	.001
Poor, n (%)	12 (0.9)	20 (3.6)	12 (1.7)	1 (1.2)	<.001	.99	.34
Very poor, n (%)	0 (0)	1 (0.2)	0 (0)	0 (0)	.29	–	.99
PA pressure, median [IQR]	25 [25]	25 [25–40]	25 [25]	25 [25]	<.001	.07	.005
Normal, n (%)	1145 (83.9)	374 (67.0)	607 (88.1)	69 (80.2)	<.001	.04	.01
Moderately increased, n (%)	157 (11.5)	122 (21.9)	48 (7.0)	10 (11.6)	<.001	.12	.03
Severely increased, n (%)	64 (4.6)	62 (11.1)	34 (4.9)	7 (8.1)	<.001	.20	.41
Prior cardiac surgery, n (%)	93 (6.8)	65 (11.6)	44 (6.4)	6 (7.0)	<.001	.82	.20
Logistic EuroSCORE, median [IQR]	3.4 [2.1–6.0]	5.3 [3.2–8.8]	2.9 [1.8–5.5]	5.5 [4.2–7.7]	<.001	<.001	.49

Abbreviations: IQR, interquartile range; LV, left ventricular; LVEF, left ventricular ejection fraction; MIDVS, minimally invasive double valve surgery; MIMVS, minimally invasive mitral valve surgery; MV, mitral valve; PA, pulmonary artery; SD, standard deviation; SDV, sternotomy double valve; SMV, sternotomy mitral valve; TV, tricuspid valve.

Table 2.

Procedural characteristics of all unmatched subgroups

	Sternotomy		Minimally invasive		p‐value
	MV N = 1365	MV + TV N = 558	MV N = 689	MV + TV N = 86	SMV vs. SDV	MIMVS vs. MIDVS	SDV vs. MIDVS
Type of procedure
MV repair, n (%)	1046 (76.6)	425 (76.2)	511 (74.2)	54 (62.8)	.83	.03	.01
MV replacement, n (%)	319 (23.4)	133 (23.8)	178 (25.8)	32 (37.2)	.83	.03	.01
TV reconstruction, n (%)	‐	554 (99.3)	‐	84 (97.7)	‐	‐	.19
TV replacement, n (%)	‐	4 (0.7)	‐	2 (2.3)	‐	‐	.19
Atrial septal closure, n (%)	37 (2.7)	14 (2.5)	20 (2.9)	5 (5.8)	.80	.13	.16
Rhythm surgery, n (%)	216 (15.8)	170 (30.5)	83 (12.0)	20 (23.3)	<.001	.01	.17
Type of prosthesis MV
Mechanical, n (%)	161 (50.6)	71 (53.4)	76 (42.7)	7 (21.9)	.59	.03	.001
Biological, n (%)	154 (48.4)	61 (45.9)	102 (57.3)	25 (78.1)	.62	.03	.001
Unknown, n (%)	3 (0.9)	1 (0.8)	0 (0)	0 (0)	.99	‐	.99
Repair rate
Primary surgery, n(%)	(1011/1272) 79.5%	(401/493) 81.3%	(499/645) 77.4%	(51/80) 63.7%	.38	.01	<.001
Redo surgery, n(%)	(35/93) 37.6%	(24/65) 36.9%	(12/44) 27.3%	(3/6) 50%	.92	.35	.67

Abbreviations: MIDVS, minimally invasive double valve surgery; MIMVS, minimally invasive mitral valve surgery; MV, mitral valve; SDV, sternotomy double valve; SMV, sternotomy mitral valve; TV, tricuspid valve.

Propensity score matching was performed for covariates that differed at baseline, i.e. age, gender and diabetes. A total of 79 pairs were matched in a 1:1 ratio. After matching, the two groups were comparable for all confounders (standardized mean difference <0.10); Table 3; Figure S1).

Table 3.

Baseline characteristics of the matched subgroups

	MIDVS	MIMVS	Prematching	Postmatching	MIDVS	SDV	Prematching	Postmatching
	N = 79	N = 79	SMD	SMD	N = 75	N = 75	SMD	SMD
Age, median [IQR] a , b	73.0 [67.0–76.0]	73.0 [67.0–76.0]	‐0.71	0.01	73.0 [67.0–77.0]	72.0 [67.0–77.0]	−0.41	0.05
Male sex, n (%) a , b	30 (38.0)	32 (40.5)	−0.42	−0.05	33 (44.0)	37 (49.3)	0.22	0.09
Diabetes, n (%) a	2 (2.5)	3 (3.8)	−0.35	0.07	9 (12.0)	8 (10.7)	−0.08	−0.05
Chronic lung disease, n (%)	7 (8.9)	7 (8.9)	−0.24	0.00	8 (10.7)	9 (12.0)	0.12	0.04
Extracardiac arteriopathy, n (%) b	1 (1.3)	1 (1.3)	0.04	0.00	1 (1.3)	1 (1.3)	0.17	0.00
Active endocarditis, n (%)	0 (0)	0 (0)	0.05	0.00	0 (0)	0 (0)	–	–
Recent myocardial infarction, n (%)	0 (0)	0 (0)	0.07	0.00	0 (0)	1 (1.3)	0.13	0.10
Serum creatinine (>200 μm/L)	0 (0)	0 (0)	0.05	0.00	0 (0)	1 (1.3)	0.14	0.10
LVEF, median [IQR] b	55 [55]	55 [55]	0.14	−0.04	55 [48–55]	55 [50–55]	−0.28	−0.01
Good, n (%)	61 (77.2)	64 (81.0)			53 (70.7)	46 (61.3)
Moderate, n (%)	17 (21.5)	13 (16.5)			21 (28.0)	27 (36.0)
Poor, n (%)	1 (1.3)	2 (2.5)			1 (1.3)	2 (2.7)
PA pressure, median [IQR] b	25 [25]	25 [25]	−0.28	0.01	25 [25]	25 [25–35]	0.18	0.04
Normal, n (%)	64 (81.0)	63 (79.7)			58 (77.3)	53 (70.7)
Moderately increased, n (%)	8 (10.1)	9 (11.4)			10 (13.3)	15 (20.0)
Severely increased, n (%)	7 (8.9)	7 (8.9)			7 (9.3)	7 (9.3)
Prior cardiac procedures			−0.02	−0.05			0.15	0.00
Prior cardiac surgery, n (%)	5 (6.3)	4 (5.1)			6 (8.0)	6 (8.0)
Logistic EuroSCORE, median [IQR]	5.2 [4.2–7.4]	5.5 [3.2–8.0]	−0.61	−0.01	5.4 [4.2–7.6]	5.9 [3.8–8.6]	0.07	0.02

Abbreviations: IQR, interquartile range; LV, left ventricular; LVEF, left ventricular ejection fraction; MIDVS, minimally invasive double valve surgery; MIMVS, minimally invasive mitral valve surgery; MV, mitral valve; PA, pulmonary artery; SD, standard deviation; SDV, sternotomy double valve; SMV, sternotomy mitral valve; TV, tricuspid valve.

^a Matching covariates MIDVS to MIMVS.

^b Matching covariates MIDVS to SDV.

In the matched population no significant difference in 30‐day and 120‐day mortality was found (Table 4). The higher incidence of new onset arrhythmia and higher need for re‐exploration of bleeding for MIDVS persisted after matching.

Table 4.

Postoperative complications of the matched subgroups

	MIDVS	MIMVS	p‐value	MIDVS	SDV	p‐value
	N = 79	N = 79		N = 75	N = 75
30‐day mortality, n (%)	0 (0)	0 (0)	‐	0 (0)	1 (1.3)	.99
120‐day mortality, n (%)	1 (1.3)	0 (0)	.99	1 (1.3)	2 (2.7)	.71
Hospital stay in days, median [IQR]	8 [5–11]	7 [5–8]	.08	9 [6–13]	7 [6–11]	.04
Perioperative myocardial infarction, n (%)	0 (0)	0 (0)	‐	0 (0)	0 (0)	‐
Pneumonia, n (%)	1 (1.3)	3 (3.8)	.62	1 (1.3)	3 (4.0)	.62
Urinary tract infection, n (%)	1 (1.3)	0 (0)	.99	0 (0)	0 (0)	‐
Reintubation due to respiratory insufficiency, n (%)	0 (0)	0 (0)	‐	0 (0)	0 (0)	‐
Prolonged intubation (>24 h), n (%)	2 (2.5)	1 (1.3)	.99	2 (2.7)	1 (1.3)	.99
Readmission to ICU, n (%)	1 (1.3)	1 (1.3)	.99	1 (1.3)	1 (1.3)	.99
Stroke, n (%)	0 (0)	1 (1.3)	.99	0 (0)	1 (1.3)	.99
Stroke with neurological deficit, n (%)	0 (0)	1 (1.3)	.99	0 (0)	1 (1.3)	.99
Stroke without neurological deficit, n (%)	0 (0)	0 (0)	‐	0 (0)	0 (0)	‐
Kidney failure, n (%)	0 (0)	0 (0)	‐	1 (1.3)	2 (2.7)	.99
Gastro‐intestinal complications, n (%)	1 (1.3)	0 (0)	.99	1 (1.3)	0 (0)	.99
Vascular complications, n (%)	1 (1.3)	0 (0)	.99	1 (1.3)	0 (0)	.99
New‐onset arrhythmia, n (%)	35 (44.3)	13 (16.5)	<.001	33 (44.0)	27 (36.0)	.32
Deep sternal wound infection, n (%)	0 (0)	0 (0)	‐	0 (0)	0 (0)	‐
Re‐exploration (within 30 days), n (%)	12 (15.2)	3 (3.8)	.02	9 (12.0)	5 (6.5)	.26
Reintervention, n (%)	5 (6.3)	6 (7.6)	.76	5 (6.7)	3 (4.0)	.72
Reintervention in primary mitral valve repair, n (%)	5 (9.8)	6 (10.9)	.85	5 (10.9)	3 (4.9)	.29

Abbreviations: ICU, intensive care unit; IQR, interquartile range; MIDVS, minimally invasive double valve surgery; MIMVS, minimally invasive mitral valve surgery; SDV, sternotomy double valve.

3.2. Minimally invasive double valve surgery compared to sternotomy double valve surgery

Patient referred for MIDVS were older (73 [67–77] vs. 70 [62–75]; p < .001) compared to patients undergoing SDV (Table 1). A higher incidence of moderate LV function and moderately increased pulmonary artery pressures were observed in sternotomy patients. No significant difference in logistic EuroSCORE were observed (SDV 5.3 [3.2–8.8] vs. MIDVS 5.5 [4.2–7.7]; p = .49). In the SDV group more MVr was performed (n = 425 [76.2] vs. n = 54 [62.8]; p = .01) resulting in a higher repair rate (81.3% vs. 63.7%; p < .001) (Table 2). No difference in 30‐day mortality (n = 0 [0] vs. n = 13 [2.3]; p = .23) and 120‐day mortality (n = 20 [3.5] vs. n = 1 [1.2]; p = .34) were observed (Table S1). Furthermore, no difference in postoperative complications, besides a higher incidence of re‐exploration for bleeding (12 [14.0] vs. n = 38 [6.8], p = .02) was observed in the MIDVS group (Table S1). Five‐year survival rate was 95.3% in the MIDVS group, compared with 90.1% in the SDV group (Log‐rank p = .12). Kaplan–Meier curves are shown in Figure 1A.

Figure 1.

Kaplan–Meier survival analysis for the total cohort and for the matched cohort. MIDVS, minimally invasive double valve surgery; MIMVS, minimally invasive mitral valve surgery; SDV, sternotomy double valve.

Propensity score matching was performed with parameters that differed at baseline, i.e. age, gender, left ventricular ejection fraction and pulmonary artery pressure. A total of 75 pairs were matched in a 1:1 ratio. After matching, the 2 groups were comparable for all confounders (standardized mean difference <0.10); Table 3 and Figure S2).

After matching no difference in 30‐day and 120‐day mortality were found and the postoperative complication rates were similar between the groups. The median length of hospital stay was longer in the MIDVS group (9 days [6–13] vs. 7 days [6–11]; p = .04) in the SDV group. (Table 4) Kaplan–Meier analyses in the matched population showed no significant difference between MIDVS and SDV (94.7% vs. 93.3%, Log‐Rank p = .62 Figure 1B).

4. DISCUSSION

The current study investigated the safety and efficacy of minimally invasive double atrioventricular valve surgery. As such, the study aim was divided into two research questions: does the addition of TV surgery add risk to a MIMVS procedure? And second, are there differences in outcomes between approaches (i.e., sternotomy or minimally invasive) for double atrioventricular valve surgery? Due to the relatively low incidence of double valve surgery through MIS, a multi‐centre nationwide registry was realized, reflecting real‐world outcomes. The main findings of the current study are that concomitant TV surgery did not increase the risk of mortality as a concomitant surgical procedure, and that there were no differences in terms of safety and efficacy between MIS and sternotomy for surgical correction of double atrioventricular valve disease, despite a higher incidence of re‐exploration for bleeding and new‐onset arrhythmias.

Mitral valve disease, right ventricular dysfunction and secondary TR are inseparably related. Although MV disease can be considered the culprit of secondary TR in many cases, the mere surgical resolution of MV disease (either MR or MS), does not necessarily eliminate the risk of future TR progression. Even in cases of mild TR during primary isolated MV surgery, TR is reported to progress to a severe class, also in case of a durable MV repair or replacement. ¹² However, outcomes of TR progression vary in the literature, and might also be related to the etiology of initial MR. ¹³ , ¹⁴ Still, re‐operative surgery for severe symptomatic isolated TR carries significant morbidity and mortality rates, ¹⁵ with a potentially prohibitive surgical risk, resulting in reduced survival of this patient group. Especially the latter observation has lowered the threshold for concomitant TV surgery during primary MV surgery. A recent robust randomized clinical trial by the Cardiothoracic Surgical Trials Network (CTSN), compared concomitant TV repair during the index procedure to isolated MV surgery in patients with degenerative MV disease and moderate TR or mild‐to‐moderate TR and TA dilatation (>40 mm). ¹⁶ Although 2‐year mortality did not differ between groups, there was a significant difference in the composite endpoint consisting of mortality, reoperation and extensive progression of TR, favouring a concomitant procedure. Of note, the observed difference was mainly driven by a marked reduction in TR progression in the concomitant group. Also, not incorporated in the composite outcome, permanent pacemaker implantation rate was surprisingly high in the concomitant group (14.1% vs. 2.5%), which is known to influence progression of heart failure and long‐term survival. ¹⁷ , ¹⁸ Still, with the expected increase in MV disease and the lowering of the threshold for indication of concomitant TV repair, a significant rise in double atrioventricular valve surgery in the near future is expected. An important‐but yet to be elucidated–observation in the current study was a reduction in repair rate in the concomitant MIMVS group (77.4% vs. 63.7%). Unfortunately, the NHR database does not include MV disease entity (i.e. MS or MR) and does not differentiate between MR aetiologies, complicating adequate interpretation of repair rate as an outcome. Another important finding was the increased rate of re‐explorations in the group undergoing double valve surgery, which could be explained by the additional incision and suture line in the right atrium. Still, derived from the results of the current study, surgeons should not be apprehended by a potentially elevated surgical risk to performed additional TV surgery during the index procedure, as morbidity and mortality rates and 120‐day survival were equal in both the unmatched and matched groups.

When an indication for concomitant TV and MV surgery is set by the heart team, the question arises which approach (i.e., through sternotomy or MIS) is more appropriate in this setting. As the mere addition of TV surgery inherently leads to longer cardiopulmonary bypass (CPB) and aortic cross‐clamping (ACC) times, this is an issue worth considering, particularly in the case of MIS. Indeed, MIS is perceived to be technically more demanding than valvular surgery performed through sternotomy, resulting in prolonged CPB and ACC times, especially in the early stages of program initiation. ¹⁹ Until now, only high‐volume expert centers have evaluated their results of minimally invasive double valve surgery compared to sternotomy. These single‐centre experiences demonstrated that double valve surgery can be performed as safely through MIS, without differences in operative result or morbidity and mortality rates, despite longer CPB and ACC time. ²⁰ These findings were confirmed in a recent meta‐analysis of observational studies. ²¹ Interestingly, the study by Akin et al. evaluated long‐term survival as well, demonstrating no differences in long‐term follow‐up between both approaches. ²⁰ Still, it remains difficult to extrapolate these findings to real‐world practice, as these studies comprised (expert) single‐centre experiences. Subsequently, the current study was initiated with data from a nationwide registry, allowing for an adequate multi‐centre interpretation of surgical results. We found no differences in terms of short‐term mortality and 5‐year survival in the unmatched and matched groups. Although there were no differences in terms of complications after matching either, a significant longer duration of hospitalization in the minimally invasive group was observed, in accordance with prior results of this nationwide registry for isolated MV surgery. ⁹ Whether this remarkable difference is related to the surgical approach or to interhospital discharge protocol differences, remains to be elucidated. Based on these findings, a MIS approach is as safe and effective as double valve surgery through sternotomy, confirmed by the excellent 5‐year survival rates of 95.3% and 90.1% in the unmatched population and 94.7% versus 93.3% in the matched population, respectively.

Of note, the current findings should not be interpreted as an unambiguous decision for MIS in all patients with an indication for double valve surgery. MIS in general, and with the use of peripheral CPB in particular, requires adequate patient selection to, amongst others, evaluate the presence of extensive peripheral arterial disease. Subsequently, we advocate for a standardized process of preoperative planning, specifically minimizing the peri‐operative risk associated with MIS.

4.1. Limitations

Although the multicentre character of the current registry allows for a real‐world evaluation of outcomes of atrioventricular double valve surgery, several limitations should be addressed. First, only a modest number of patients undergoing combined MV and TV surgery were operated through MIS. Also, as patients could not be traced back to the participating centers, a possible inclusion bias exists as hospital and operator volume remain unclear. Furthermore, although propensity score matching was performed to correct for known confounders, this method cannot balance unknown confounders, potentially leaving some form of selection bias. Finally, given its retrospective nature, the registry was subjected to some missing data, such as CPB and ACC times. Although we could not assess differences in the surgical times, the expected increased CPB and ACC time in the concomitant groups did not result in clinical sequelae, as demonstrated by our data. As the registry does not contain any information on the actual approach of the mitral valve (left atriotomy or transseptal approach) for patients operated through sternotomy, the higher incidence of new‐onset arrhythmia compared with sternotomy should be interpreted with care.

5. CONCLUSION

The current multicenter nationwide study presented the outcomes of minimally invasive double atrioventricular valve surgery. No differences in short‐term mortality and 5‐year survival compared to both minimally mitral valve surgery and double valve surgery through sternotomy were found. A higher incidence of re‐exploration for bleeding and postoperative arrhythmia and longer hospital stay was observed in the minimally invasive double valve group compared to minimally invasive single valve surgery. The patient selection remains crucial with the expected increase in double valve surgery.

CONFLICT OF INTEREST

Peyman Sardari Nia has a consultancy agreement with Neochord Inc, Edwards Lifesciences, Fuijfilm medical and is the inventor of a mitral valve simulator that is commercialized through a start‐up (Ma‐trac) of Maastricht University Medical Centre+.

ETHICS STATEMENT

No individual centre institutional review board approval was necessary for this retrospective study. This study is in line with the institution's ethical policies and standards.

Supporting information

Supplementary information.

APPENDIX 1. Cardiothoracic Surgery Registration Committee of the Netherlands Heart Registration

Dr. S. Bramer Amphia

Dr. W.J.P. van Boven Amsterdam UMC, locatie AMC

Dr. A.B.A. Vonk Amsterdam UMC, locatie VUmc

Drs. B.M.J.A. Koene Catharina Ziekenhuis

Dr. J.A. Bekkers Erasmus MC

Dr. G.J.F. Hoohenkerk HagaZiekenhuis

Dr. A.L.P. Markou Isala

Drs. A. de Weger Leids Universitair Medisch Centrum

Dr. P. Segers Maastricht UMC+

Drs. F. Porta Medisch Centrum Leeuwarden

Dr. R.G.H. Speekenbrink Medisch Spectrum Twente

Dr. W. Stooker OLVG

Drs. W.W.L. Li Radboudumc

Drs. E.J. Daeter St. Antonius Ziekenhuis

Dr. N.P. van der Kaaij UMC Utrecht

Dr. Y.L. Douglas Universitair Medisch Centrum Groningen

Olsthoorn JR, Heuts S, Houterman S, et al. Does concomitant tricuspid valve surgery increase the risks of minimally invasive mitral valve surgery? A multicentre comparison based on data from The Netherlands Heart Registration. J Card Surg. 2022;37:4362‐4370. 10.1111/jocs.17004