Biomarkers of inflammation and coagulation after minimally invasive mitral valve surgery: a prospective comparison to conventional surgery

STRUCTURED ABSTRACT

Objectives

Minimally invasive cardiac surgery techniques are increasingly used but have longer cardiopulmonary bypass time, which may increase inflammatory response and negatively affect coagulation. Our aim was to compare biomarkers of inflammation and coagulation as well as transfusion rates after minimally invasive mitral valve repair and mitral valve surgery using conventional sternotomy.

Design

A prospective non-randomized study was performed enrolling 71 patients undergoing mitral valve surgery (35 right mini-thoracotomy and 36 conventional sternotomy procedures). Blood samples were collected pre- and postoperatively to assess inflammatory response. Thromboelastometry (ROTEM) was performed to assess coagulation, and transfusion rates were monitored.

Results

The minimally invasive group had longer cardiopulmonary bypass times compared to the sternotomy group: 127 min ([115–146] vs 79 min [65–112], p < 0.001) and were cooled to a lower temperature during cardiopulmonary bypass, 34 °C vs 36 °C (p = 0.04). IL-6 was lower in the minimally invasive group compared to the conventional sternotomy group when measured at the end of the surgical procedure, (38 [23–69] vs 61[41–139], p = 0.008), but no differences were found at postoperative day 1 or postoperative day 3. The transfusion rate was lower in the minimally invasive group (14%) compared to full sternotomy (35%, p = 0.04) and the chest tube output was reduced, (395 ml [190-705] vs 570 ml [400–1040], p = 0.04).

Conclusions

Our data showed that despite the longer use of extra corporal circulation during surgery, minimally invasive mitral valve repair is associated with reduced inflammatory response, lower rates of transfusion, and reduced chest tube output.

Keywords:

• Minimally invasive surgery
• coagulopathy
• mitral valve repair
• inflammatory response
• ROTEM

Background

Minimally invasive techniques using a mini-thoracotomy are increasingly used in mitral valve (MV) repair with similar surgical outcomes as conventional sternotomy [1–3]. The potential benefits of minimally invasive mitral valve surgery (MIMVS) are improved cosmetics [4], pain reduction [5], and a faster return to normal activity [6]. However, minimally invasive techniques usually require longer cardiopulmonary bypass times, which may diminish the benefits of reduced inflammation resulting from less extensive surgical trauma. It has been shown that prolonged cardiopulmonary bypass (CPB) run time may increase the inflammatory response and affect coagulation negatively [3,7]. Furthermore, transfusion of allogenic red blood cells (RBCs) has been debated as a risk factor for adverse outcome after cardiac surgery [8–10], also resulting in significant costs [11,12].

There also is data suggesting that MIMVS is associated with reduced blood loss, fewer transfusions, and faster recovery compared to the standard approach [13,14]. There are, however, few studies comparing the relationship between inflammation, coagulation, and bleeding between MIMVS and MV repair by full sternotomy.

Our aim was to compare inflammatory and coagulation biomarkers as well as ROTEM® variables and transfusion rates between MIMVS and MV surgery using a conventional sternotomy.

Patients and methods

This was a prospective non-randomized study including 71 patients undergoing isolated mitral valve surgery (35 right mini-thoracotomy and 36 conventional sternotomy procedures). Patients undergoing planned MV repair were recruited and provided written informed consent. Patient undergoing closure of the atrial appendage and or a persistent foramen ovale were included. Exclusion criteria were other concomitant procedures and redo-surgery. The study protocol was reviewed and approved by the Swedish Ethical Review Agency (ref ID: 2017/696).

Primary outcome of the study was the release of specific inflammatory biomarkers (IL-6, CRP, and procalcitonin) and coagulation biomarkers (ROTEM, fibrinogen, platelets, APTT, PT INR). Secondary outcome was the need for any transfusion, reoperation for bleeding, and postoperative bleeding volume.

Minimally invasive mitral valve surgery was introduced at our department in 2010 and has been used for patients undergoing elective MV surgery according to surgeons’ preference. Preoperative screening included angiographic imaging of the aorto-iliac vessels to rule out severe atherosclerotic disease. Relative exclusion criteria were advanced age, obesity, renal failure, previous cardiac surgery, extensive annular calcifications, and difficult vascular access. Patients in the MIMVS group were intubated with a single or double lumen endotracheal tube. Surgical access was obtained through a right anterior thoracotomy at the fourth intercostal space via a 5–6 cm skin incision. Cardiopulmonary bypass was installed through cannulation of the femoral artery and vein via guidewires under the guidance of transesophageal echocardiography. Cardioplegic arrest was achieved by antegrade administration of 1000 ml cold crystalloid cardioplegia (Custodiol® HTK Solution, Essential Pharmaceuticals, Durham, USA) in the aortic root after aortic cross-clamping with a Chitwood clamp. Vacuum-assisted venous drainage was utilized when necessary (max −40 mmHg). Ultrafiltration was utilized during CPB to remove excessive volume related to Custodiol® infusion.

The conventional group had median sternotomy with standard cannulation of the distal ascending aorta and bicaval cannulation. Cardioplegic arrest was achieved with intermittent cold blood cardioplegia (antegrade and retrograde in combination). MV repair techniques were similar in both groups with neochords and leaflet preservation.

Blood samples were collected at four occasions pre- and postoperatively (T₀ - start of surgery skin incision; T₁ - after wound closure; T₃ – post-operative day 1; T₄ - post-operative day 3) to assess inflammatory response and standard biomarkers of coagulation. Thromboelastometry was performed at T₀ and T₁ to assess coagulation and, finally, transfusion rates were monitored until discharge.

All blood samples were analyzed at the Dept. of Clinical Chemistry, Division of Laboratory Medicine, Skåne University Hospital, Lund, Sweden. Procalcitonin levels were analyzed using Atellica IM BRAHMS PCT (Thermo Fisher Scientific Inc, Waltham, Massachusetts, USA). IL-6 levels were analyzed using Immulite® 1000 (Siemens, München, Germany). The ROTEM variables analyzed were clotting time (CT) measured in seconds (s) in EXTEM, INTEM and HEPTEM, and maximum clotting firmness (MCF) in millimetres (mm) in EXTEM and FIBTEM. ROTEM was analyzed using ROTEM Delta 4000 (Tem Innovations GmbH, Münich, Germany).

Statistical analysis

Categorical variables were expressed as numbers and percentages. Continuous variables were expressed as mean (±SD) when normally distributed, otherwise as median [IQR]. Categorical variables were compared by use of Chi-squared or Fisher exact test when the expected value in any of the cells of the contingency table was below 5. Continuous data were compared by groups with the use of Student’s t-test if normally distributed or by the Mann–Whitney test.

The analyses were made using SPSS software 28.0.0.0 (IBM, Armonk, NY, USA). p-values of <0.05 were considered statistically significant.

Results

Patient and operative characteristics

Patient characteristics are presented in Table 1. The MIMVS population had lower mean age, (55 ± 11 vs 62 ± 12, p = 0.02), a lower EuroScore II, (0.78 ± 0.36 vs 1.11 ± 0.72, p = 0.019), and a lower rate of preoperative treatment with aspirin (0 (0%) vs 8 (22.%), p = 0.005). The operative characteristics are presented in Table 2. One patient (3%) was converted to sternotomy. One patient (3%) in the sternotomy group scheduled for valve repair eventually had valve replacement due to repair failure. The minimally invasive patients had longer cardiopulmonary bypass times (127 min [115–146] vs 79 min [65–112], p < 0.001), and were cooled to a lower body temperature during cardiopulmonary bypass (34 °C vs 36 °C, p = 0.04), when compared to the sternotomy group. There was no in-hospital mortality in either group.

Table 1. Patient characteristics in minimally invasive surgery group vs conventional technique group.

Characteristic	MIMVS n = 35 (%)	Conventional group n = 36 (%)	p value
Age (years)	56 ± 11	62 ± 12	0.02
Female gender	3 (9)	8 (22)	0.11
Height (cm)	178 ± 8	177 ± 10	0.67
Weight (kg)	76 ± 9	83 ± 16	0.07
Hypertension	11 (32)	17 (47)	0.20
Diabetes	1 (3)	1 (3)	0.98
COPD	0	1 (3)	0.32
Creatinine	80 ± 13	84 ± 16	0.21
Preoperative LVEF 30–50%	2 (6)	2 (6)	1
Preoperative LVEF <30%	0	0
NYHA Class III-IV	6 (17)	1 (3)	0.11
Pulmonary hypertension (>60mmHg)	6 (17)	2 (6)	0.26
Previous AFib	2 (6)	6 (17)	0.15
Peripheral vascular disease	0	0	1
Previous cardiac surgery	0	0	1
EuroScore 2	0.78 ± 0.36	1.11 ± 0.72	0.019
Preoperative Aspirin	0	8 (22)	0.005

Values were expressed as number and percentage (%) or as mean ± standard deviation. ASA: acetyl salicylic acid.

Table 2. Operative characteristics in minimally invasive surgery group vs conventional technique group.

Characteristic	MIMVS n = 35 (%)	Conventional group n = 36 (%)	p value
CPB time (min)	127 [115–146]	79 [65–112]	<0.001
Aortic cross clamp (min)	76 [58–82]	56 [42–90]	0.07
Lowest temperature (°C)	34 [33–36]	36 [35–36]	0.04
Conversion to sternotomy	1 (3)		NA
Mitral valve replacement	0	1 (3	0.43
Post operative MR
None	28 (80)	24 (67)
Trace	6 (17)	8 (22)
Grade 1	0	1 (3)

Values were expressed as number and percentage (%) or as median [IQR]. CPB: cardiopulmonary bypass; MR: mitral regurgitation.

Table 3. Postoperative bleeding/transfusion outcomes in minimally invasive surgery group vs conventional technique group.

Variable	MIMVS n = 35 (%)	Conventional group n = 36 (%)	p value
Bleeding/drain volume (ml)	395 [190–705]	570 [400–1040]	0.04
Revision bleeding	1 (3)	1 (3)	0.98
Any transfusion	5 (14)	12 (35)	0.04
Packed RBC	3 (9)*	8 (24)	0.09
Plasma	1 (3)	6 (17)	0.11
Platelets	3 (9)	7 (19)	0.31

Values were expressed as number and percentage (%), or as median [IQR]. RBC: packed red blood cells.

Inflammatory reaction

IL-6 was lower in the minimally invasive group compared to the conventional sternotomy group when measured at the end of the surgical procedure (38 ng/L [23–69] vs 61 ng/L [41–139], p = 0.01), but no statistically significant differences were demonstrated preoperatively or at postoperative day 1 or 3. Procalcitonin was higher in the minimally invasive group compared to the conventional sternotomy group on postoperative day 1 (0.50 µg/L [0.35–1.00] vs 0.30 µg/L [0.16–0.49], p = 0.008), but no significant differences were observed preoperatively, at the end of the surgical procedure or on postoperative day 3. CRP levels were reduced in the minimally invasive group compared to the conventional sternotomy group when measured on postoperative day 1 (38 mg/L [32–48] vs 61 mg/L [50–78], p < 0.001), but no significant differences were found preoperatively, at the end of the surgical procedure, or on postoperative day 3. Inflammation markers are presented in Figure 1. There were no differences in leukocyte counts between the two groups at any time.

Figure 1. Inflammatory response MIMVS vs CS.

Coagulopathy and transfusions

Antithrombin levels were significantly higher in the minimally invasive group both preoperatively (0.90 ± 0.10 IU/L vs 0.85 ± 0.11 IU/L, p = 0.03) and at the end of the surgical procedure (0.76 (±0.10) IU/L vs 0.71 (±0.13) IU/L, p = 0.02), but no significant differences could be detected at postoperative day 1 or 3. No significant differences were observed in levels of D-dimer, prothrombin time internationalized normalized ratio (PT-INR), activated partial thromboplastin time (APT-T), or platelets at any time point. ROTEM clotting times (CTs) did not differ at any point in time between the groups, neither did maximum clotting firmness (MCF) in EXTEM nor FIBTEM. ROTEM is presented in Figure 2. Transfusion data is presented i Table 3. The number of patients receiving transfusions before discharge was lower after minimally invasive mitral surgery when compared to full sternotomy (14% vs 35%, p = 0.04). Chest tube output on postoperative day 1 was lower in the minimally invasive group compared to the sternotomy group (395 ml [190–705] vs 570 ml [400–1040], p = 0.04).

Figure 2. ROTEM.

Discussion

Our study demonstrates that, despite longer duration of CPB, MIMVS is associated with reduced inflammatory response and a lower transfusion rate. Furthermore, there were no differences in coagulation activation detectable by ROTEM or routine lab analyses.

Minimally invasive MV repair is an established and reproducible procedure, but still, the adoption of this technique is rather limited in the cardiothoracic community. This is partly due to the relatively long learning curve [15], which usually means longer time on CPB. Prolonged time on CPB has negative effects on both coagulation and inflammation. CPB has been shown to activate proinflammatory cytokines such as IL-8 and IL-10, complement factors, and neutrophile leukocytes [16]. The inflammatory response has been shown to increase the risk of atrial fibrillation [17], kidney dysfunction [18], and respiratory problems [16,19,20,21]. The negative effects of CPB on coagulation have been thoroughly studied and are multifactorial. CPB induces a consumptive coagulopathy, which leads to a decrease in coagulation factor V, VII, VIII, and IX levels as well as a decrease in fibrinogen and platelet number and function [22].

IL-6 and CRP are well-studied inflammatory response markers and have been linked to worse outcomes in cardiac surgery patients [17,19,20]. In this study, we could show that MIMVS reduces IL-6 and CRP compared to conventional sternotomy, which is in line with the study by Paparella et al. who compared conventional sternotomy with minimally invasive aortic valve replacement and mitral valve repair. Both these studies further confirm the conclusion by Asimakopoulos [16] that increased IL-6 and CRP are more a product of surgical trauma rather than CPB.

Procalcitonin is another marker for inflammatory response, most commonly used to detect bacterial infections [23], but also is elevated in non-infectious inflammatory responses [24–26] and has been associated with poorer outcomes after cardiac surgery [19,27]. Kilger et al. [28] compared procalcitonin in patients undergoing coronary bypass with conventional sternotomy with minimally invasive off-pump coronary bypass (MIDCAB) and demonstrated lower procalcitonin in the MIDCAB group. Our results contrast with their findings, with higher procalcitonin in the MIMVS group on the first postoperative day. A possible explanation could be the increased CPB time in the MIMVS. This association has been supported by several studies [29–31], although the causality of CPB-effects on procalcitonin has been debated [32,33].

ROTEM is widely used to monitor coagulation disturbances in both elective and acute settings. To the best of our knowledge, this is the first study comparing ROTEM in MIMVS to CS. Despite longer CPB time in the MIMVS group, this study could not identify any negative effects on ROTEM variables or in routine laboratory coagulation tests. This indicates that coagulation activation is similar between groups despite the longer CPB time in MIMVS. Consequently, this suggests that the reduced bleeding volume and transfusion requirement in MIMVS is an indicator of reduced surgical trauma rather than alterations of the coagulation system.

Our data demonstrated a reduced number of transfusions and lower chest tube output in patients undergoing minimally invasive surgery. A reduction in postoperative hemorrhage and transfusion requirements has been suggested as a potential advantage of minimally invasive valve surgery. This benefit is important given the significant morbidity and mortality associated with transfusions and re-exploration for bleeding [10]. In a consecutive series of 41 patients undergoing either MIMVS (n = 21) or CS (n = 20), Glower et al. [6] demonstrated no significant difference in chest tube output or transfusion requirements despite longer CPB times in the minimally invasive procedure.

Limitations

The major limitation of our study is the lack of randomization. Our groups had a different preoperative clinical profile and the conventional group was older and had more aspirin, but randomization was not considered feasible and ethical during study conception because patients’ preference and surgeon experience play a major role in the choice of the surgical approach. Hemofiltration was only used in MIMVS group and could potentially influence inflammatory markers in either direction. Being an exploratory analysis, a formal sample size calculation was not performed. To reduce the risk of type 2 errors, we based our sample size on previous studies of our group with similar endpoints.

Conclusions

This study demonstrated that although MIMVS is associated with a longer duration of CPB, the minimally invasive technique reduced inflammatory activation, had no detrimental impact on coagulation activation, and was associated with a reduced need for transfusions compared to standard sternotomy. Our data offer further evidence supporting the adoption of minimally invasive approaches.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Not possible due to limitations in ethical approval.

Additional information

Funding

This study was supported with grants from the Ollie and Elof Ericsson's Foundation for Scientific Research.