Does a minimally invasive approach result in better pulmonary function postoperatively when compared with median sternotomy for coronary artery bypass graft?

Abstract

A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was ‘Does a minimally invasive approach result in better pulmonary function postoperatively when compared with median sternotomy for coronary artery bypass graft?’. Procedures such as limited sternotomy and minimally invasive direct coronary artery bypass (MIDCAB) though a minithoracotomy were regarded as minimally invasive. Overall, 681 papers were found, of which eight represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, evidence level, relevant outcomes and results of these papers were tabulated. Three randomized, controlled trials (RCT) were included: One study suggested that ministernotomy dividing the corpus sterni (n = 50) offers no advantage over standard sternotomy (n = 50) during the first 10 postoperative days. Two further studies reported on minithoracotomy: one trial presented data suggesting that minithoracotomy (n = 21) is as safe as standard sternotomy with (n = 18) or without (n = 19) cardiopulmonary bypass, but without the benefit ascribed to the minimally invasive incision. A two-centre report investigated pulmonary function as a secondary outcome and claimed that minithoracotomy worsens FEV₁ and FVC. The study was not powered to detect these differences as pulmonary function data were available only for one of the centres. Five non-randomized reports were also included in this analysis: These investigated outcomes after minithoracotomy or limited sternotomy compared with standard sternotomy. Patient groups were small, involving <20 subjects per group. Non-randomized studies suggested a benefit to postoperative lung function in using thoracotomy. One of these reports included only patients with severe chronic obstructive pulmonary disease (COPD) (FEV₁ <70% of predicted) and detected benefits in selected patients undergoing MIDCAB. A further study was in agreement with the above statement in patients without COPD. MIDCAB may be more painful initially, but results in quicker recovery of lung function. Demonstrating the benefits of ministernotomy compared with the standard sternal incision was less clear. One paper demonstrates better outcomes when compared with standard sternotomy, while another reports no difference. We conclude that non-randomized studies support the hypothesis that minimally invasive coronary artery bypass benefits postoperative lung function in patients with known respiratory problems.

INTRODUCTION

A best evidence topic was constructed according to a structured protocol. This is fully described in the ICVTS [1].

THREE-PART QUESTION

In adult patients [requiring coronary artery bypass graft] does [a minimally invasive approach] result in [better postoperative pulmonary function]?

CLINICAL SCENARIO

You are a cardiac surgeon confronted with a 55-year old male patient with chronic obstructive pulmonary disease (COPD) (FEV₁: 70% of predicted) with single-vessel disease involving a critical ostial lesion of the left anterior descending (LAD) artery. You are experienced in both minimally invasive and conventional cardiopulmonary bypass grafting (CABG) techniques. Given the pulmonary impairment caused by the patient's COPD, you wonder if using the minimally invasive approach will enable a better recovery in pulmonary function for your patient.

SEARCH STRATEGY

MEDLINE 1950 to December 2012 using PubMed interface. [coronary artery bypass surgery] AND [minimal access surgical procedures/OR MIDCAB/OR median sternotomy/OR ministernotomy/OR sternotomy] AND [pulmonary function/OR respiratory function test/OR respiratory function]. A search of reference lists and related articles was conducted.

SEARCH OUTCOME

The PubMed search found 681 papers. Titles were screened and abstracts of promising articles reviewed. Studies with outcomes in terms of FEV₁, FVC and FEV₁/FVC ratio were included, but those using only the broad term ‘pulmonary complications’ were excluded. Eight studies represented the best evidence, three were randomized and five were cohort studies. One additional study identified was excluded because the article was in Japanese [2].

RESULTS

Rogers et al. [3] conducted a randomized, controlled trial across two centres in the UK and Italy. Pulmonary function was a secondary outcome and only available for UK patients. The group reported worsened mean FEV₁ and mean FVC at discharge in their thoracotomy (ThoraCAB) patients. The study was not powered for detecting differences in lung function. Results showed that ThoraCAB patients were discharged on postoperative Day 5 (interquartile range (IQR): 4–7) vs Day 6 (IQR: 4–7) for OPCAB. ThoraCAB patients had their lung function measured earlier in recovery. Requirements for postoperative analgesia were longer among ThoraCAB patients, but with a similar total administered dose (Table 1).

Table 1:

Best evidence papers

Author, date, journal and country, Study type (level of evidence [11])	Patient group	Outcomes	Key results	Comments
Rogers et al. (2012), J Thorac Cardiovasc Surg, UK and Italy [3] Open randomized controlled trial (level 2)	184 patients for elective coronary bypass grafting randomized into OPCAB group: n = 93, off-pump CABG by median sternotomy and ThoraCAB group: n = 91 (63), MIDCAB via left anterolateral thoracotomy—incision in 4th or 5th left intercostal space, with lung deflated or compressed gently where single-lung ventilation was not possible. Mean age: ThoraCAB: 63.1 years (92% male, mean BMI: 27.6) OPCAB: 66.7 years (86% male, mean BMI: 28) Pulmonary measures were secondary outcomes, for which OPCAB: n = 64 and ThoraCAB: n = 61	FEV1 at discharge, mean (SE) Mean difference ThoraCAB–OPCAB (95% CI) FVC Mean difference ThoraCAB–OPCAB (95% CI) Peak expiratory flow Mean difference ThoraCAB–OPCAB (95% CI) FEV1/FVC ratio (%) Mean difference ThoraCAB–OPCAB (95% CI)	OPCAB: 1.61 (0.09) ThoraCAB: 1.48 (0.10), −0.13 (−0.25, 0.02); P: 0.09 OPCAB: 2.12 (0.11) ThoraCAB: 1.90 (0.11), −0.025 (−0.44, −0.06); P: 0.01 OPCAB: 255 (21) ThoraCAB: 229 (21), −26 (−62,11); P: 0.17 OPCAB: 77.7 ThoraCAB: 229 (21), 1.01 (0.96, 1.07); P: 0.58	Two centres involved but pulmonary function was one of the secondary outcome measures, for which data was only available for UK patients. The study was not powered to detect differences in the secondary outcomes. This study also looked at analgesia use—37% longer in ThoraCAB, but total administered doses were similar. In conclusion, the quality-of-life data from the same trial suggests ThoraCAB is not unsafe or inferior after 12 months, despite worse mean FEV1 and FVC at discharge
Bauer et al. (2001), J Thorac Cardiovasc Surg, Germany [4] Prospective randomized study (level 2)	100 patients for elective first-time coronary artery bypass grafting were randomized into Group 1: n = 50 standard sternotomy and Group 2: n = 50 partial median ‘ministernotomy’—defined as complete splitting of corpus sterni, with an inverted-V cut into the manubrium at the second intercostal space, requiring a special retractor. Lung function measured by constant volume body plethysmography Mean age Group 1: 61.9 years Group 2: 62.9 years Male:female ratio Group 1: 39:11 Group 2: 41:9 Mean BMI Group 1: 26.9 Group 2: 26.8 Two-vessel disease Group 1: 17/50 Group 2: 13/50 Triple-vessel disease Group 1: 33/50 Group 2: 37/50	% of preoperative value VC % of preoperative value FEV1 % of preoperative value FEV1/%FVC % of preoperative value TLC	Day 4 PO Group 1: 46.9 ± 12.3 Group 2: 48.5 ± 12.6 Day 10 PO Group 1: 73.4 ± 16.2 Group 2: 71.2 ± 13.9 Day 4 PO Group 1: 49.4 ± 12.4 Group 2: 48.7 ± 12.5 Day 10 PO Group 1: 70.6 ± 9.6 Group 2: 73.7 ± 13.4 Day 4 PO Group 1: 104.9 ± 15.5 Group 2: 102.1 ± 12.3 Day 10 PO Group 1: 97.8 ± 9.1 Group 2: 103.9 ± 9.4 Day 4 PO Group 1: 57.5 ± 14.7 Group 2: 59.6 ± 13.7 Day 10 PO Group 1: 79.3 ± 9.8 Group 2: 71.8 ± 13.8	This is a detailed report. Excluded patients with BMI >35. Postoperative pain management data not collected so unclear if analgesic medication is making a difference between the two groups. Defined anaesthetic and surgical protocols used in all patients. In conclusion, the less invasive incision offers no early benefit to postoperative lung function
Guler et al. (2001), Ann Thorac Surg, Turkey [5] Prospective randomized study (level 2)	58 patients for elective isolated coronary surgery with severe COPD, randomized into three groups: Group 1: n = 18, CABG with median sternotomy and CPB; Group 2: n = 19, beating heart technique with median sternotomy but no CPB; Group 3: n = 21, MIDCAB via left anterior mini thoracotomy (precise definition of incision unclear) Mean age: Group 1: 54.05 years Group 2: 57.85 years Group 3: 52.25 years (P: 0.158) Minimal demographic information provided, but all patients had single-vessel disease, with LITA grafting to the proximal LAD artery	Postoperative FVC Postoperative FEV1/FVC Postoperative PaO2	Group 1: 54.75 ± 10.97 Group 2: 56.65 ± 9.68 Group 3: 58.15 ± 8.23 Group 1: 50.35 ± 12.65 Group 2: 53.3 ± 12.04 Group 3: 51.6 ± 12.01 Group 1: 68.15 ± 9.52 Group 2: 67.3 ± 9.9 Group 3: 65.1 ± 9.66	Three study groups allow comparison of effects of on-pump vs off-pump and sternotomy vs minimally invasive incision. Unclear if more than one surgeon was performing the operation, and little information is given about randomization process. All patients included had severe COPD [FEV1 <70%, FVC <70% and FEV1/FVC <60%]. Follow-up time was at a mean of 2 months. In conclusion, avoiding CPB when performing coronary revascularization is beneficial in patients with COPD, and in selected patients, MIDCAB can be used to avoid some unwanted effects of sternotomy and CPB
Kirali et al. (2005), J Card Surg, Turkey [6] Prospective cohort study (level 3)	17 patients for elective single-vessel revascularization using awake high thoracic epidural anaesthesia (T1–T2 level): Group A: n = 10, reversed-J inferior sternotomy—sternum divided up to second intercostal space, at which point it was cut obliquely to the left while voiding the LITA. Group B: n = 7, full sternotomy Mean age: Group A: 50.2 years Group B: 57.3 years (P: 0.051) % Male: Group A: 100% Group B: 57% Mean BMI: Group: A: 40.1 Group B: 40.8 Hypertension: Group 1: 10% Group 2: 71.4% (P: 0.018) Other CVS risk factors not statistically different between groups. 7 sternotomy operations performed first, so as to be confident if needing to convert to full sternotomy while doing newly learnt reversed-J technique	Predicted vital capacity (%) Predicted forced vital capacity (%) FEV1/FVC ratio (%)	Preoperative Group A: 94.2 ± 11.9 Group B: 102.1 ± 11.2 Postoperative Group A: 43.6 ± 13.6 Group B: 52.8 ± 12.8 Preoperative Group A: 99.4 ± 12.8 Group B: 109.5 ± 10.5 Postoperative Group A: 47.2 ± 12.3 Group B: 53.3 ± 10.5 Preoperative Group A: 77.7 ± 9.6 Group B: 81.2 ± 5.9 Postoperative Group A: 76.7 ± 8.5 Group B: 78.5 ± 2.4	This study involved a different anaesthetic protocol. Study numbers are small. Lung function was measured by spirometry 3 days before surgery and on postoperative day 3, once epidural analgesia was stopped. In conclusion, reversed-J incisions are less traumatic and offer advantages particularly intraoperatively in patients breathing spontaneously, but improved intraoperative oxygenation does not translate to improved pulmonary function in the postoperative period
Ohkado et al. (2002), JJTCVS, Japan [7] Prospective cohort study (level 3)	Patients for CABG between August 1998 and December 1999 were placed in three groups: Group M: n = 8, underwent MIDCAB—incision of 8–10 cm in 4th or 5th left intercostal space; Group O: n = 10, underwent off-pump CABG (OPCAB); Group C: n = 12, underwent conventional CABG Mean age: Group M: 69 years Group O: 67 years Group C: 62 years Male:female ratio: Group M: 5:3 Group O: 8:2 Group C: 11:1 Mean no. of grafts: Group M: 1.3 Group O: 2.1 Group C: 3.4 Group M targeted LAD artery for grafting using LITA or inferior epigastric, Group O targeted LAD and RCA and Group C targeted all coronary arteries with harvested grafts. Pulmonary function was measured using spirometry	% VC ratio at 2 weeks postoperative % FEV1 ratio at 2 weeks postoperative	Group M: 0.96 ± 0.04 Group O: 0.81 ± 0.05 Group C: 0.81 ± 0.03 Group M: 0.94 ± 0.05 Group O: 0.88 ± 0.03 Group C: 0.82 ± 0.03 (for Group M vs Group C, P < 0.05, but not for Group M vs Group O)	Exclusion criteria including any postoperative complication, is potentially quite limiting, as suggested by the sample size of n = 30 which is small for an 18-month study period. Use of three groups allows comparison between effects of minimally invasive incision and CPB is an advantage. In conclusion, MIDCAB offers improved lung function recovery over the early postoperative period than conventional CABG, and this study supports the hypothesis that it is due to the minimally invasive incision, not avoiding CPB
Lichtenberg et al. (2000), Ann Thorac Surg, USA [8] Prospective cohort (level 3)	Group A: 15 men (mean age: 65.2 ± 10.9 years) underwent MIDCAB without CPB by an anterolateral incision of 8 cm in 4th or 5th left intercostal space, involving temporary one-lung ventilation. Portable spirometer used in a best of three fashions: Group B; 15 men (mean age: 60.7 ± 9.4 years) underwent conventional CABG with CPB Mean weight: Group A: 90.9 kg Group B: 79.9 kg Mean no. of grafts Group A: 1.0 Group B: 2.7 (P ≤ 0.001)	% Predicted VC % Predicted FEV1 Pain on forced inspiration	Day 3 PO Group A: 59.7 ± 11.1 Group B: 40.6 ± 14.5 Day 5 PO Group A: 74.4 ± 10.4 Group B: 53.9 ± 12.5 Day 3 PO Group A: 56.3 ± 11.0 Group B: 42.2 ± 17.6 Day 5 PO Group A: 68.4 ± 9.0 Group B: 55.5 ± 14.8 Day 1 PO Group A: 5.5 ± 0.9 Group B: 3.6 ± 0.8 Day 3 PO Group A: 4.0 ± 1.2 Group B: 2.9 ± 0.7	Small study numbers (two groups of n = 15 each), women were excluded, as were patients with existing respiratory disease. Five days is a short follow-up period. In conclusion, MIDCAB is more painful in the early postoperative period, but allows a quicker recovery to preoperative levels of lung function
Guizilini et al. (2010), Arq Bras Cariol, Brazil [9] Prospective, non-controlled trial (level 4)	Eighteen patients with angiographic evidence of coronary artery disease, split into two groups, conventional mid-sternotomy (CMS, n = 10) and ministernotomy (MS, n = 8)—xiphoid and lower part of sternum sectioned longitudinally. Intact pleural cavity in both groups. LITA anastomosed to the anterior descending artery Mean age: CMS: 60.1 years MS: 58.5 years Male:female ratio: CMS: 9:1 MS: 7:1 BMI: CMS: 25.75 vs MS: 26.81 Lung function measured with a portable spirometer, in a best-of-three fashion	Recovery of FVC (as % of preoperative value) PaO2 (on air) as % of preoperative value Shunt (% increase)	Day 1–5 PO MS: 30.9% CMS: 22.32% Day 1 PO MS: 78.11 ± 5.0 CMS: 88.19 ± 3.29 Day 1 PO MS: 0.24 ± 0.02 CMS: 0.27 ± 0.01	Small study number. Patients with existing COPD excluded. Patients in both groups received same general anaesthesia and LITA mobilization. Spirometry performed by the same practitioner. The study also recorded pain in all enrolled patients. The study is designed to ‘eliminate the effect of SIRS induced by CPB’ [9]. In conclusion, ministernotomy allowed lung function to be better preserved and recovered relative to median sternotomy
Grossi et al. (1999), Eur J Cardiothorac Surg, USA [10] Prospective cohort (level 3)	29 patients undergoing CABG between January and July 1998, in two groups; port access (PA, n = 14) defined as 5–8 cm incision in the inframammary skin fold (on the right is single graft required) with concurrent femoral cannulation and endoaortic balloon to occlude aorta and administer cardioplegic solution, and standard sternotomy (STD, n = 15) Mean age: PA: 61.9 years STD: 68.3 (P: 0.05) PA: 76.9% male vs STD: 100% male COPD PA: 15.4% vs STD 13.3% No. of grafts needed PA: 2.33 STD: 3.33 P: 0.001	FEV FVC, FEV/FVC Pain score (1: no pain, 10: worst possible pain)	Day 1 (P < 0.02) PA: 1.59 l/s STD: 0.97 l/s Day 3 PA: 2.20 l/s STD: 1.49 l/s No statistical difference over course of whole follow-up Day 1 PA: 2.09 ± 2.32 STD: 5.7 ± 2.35 1 week PA: 0.76 ± 1.19 STD: 2.35 ± 1.69 4 weeks PA: 0.58 ± 1.72 STD: 3.43 ± 2.99	Approach chosen was surgeon's preference. Only 6 patients per group underwent pulmonary function testing. Many other outcomes were also included in the study, including markers of postoperative stress. The authors conclude that this minimally invasive approach offers less postoperative pain and better lung function and quality of life in the early postoperative period than full sternotomy, and recommend their results be substantiated by larger studies

Bauer et al. [4] studied lung function in 100 patients undergoing CABG by median sternotomy or partial ministernotomy. Reductions at Day 4 compared with preoperative values were similar between groups. By Day 10, both groups had recovered equally. The authors hypothesize that the standard sternotomy group may be receiving more analgesia to achieve the same levels of respiratory function and that a better cosmetic impact is the only advantage.

Guler et al. [5] randomized 58 elective patients with severe COPD (FEV₁ <70%) into three groups: CABG with median sternotomy and cardiopulmonary bypass, off-pump with median sternotomy and MIDCAB. They found a difference in FEV₁ at 2 months postoperatively (46.1 ± 11.9%) compared with (53.95 ± 11.65%) preoperatively in the CPB group only. The authors suggest that avoidance of CPB determines postoperative respiratory function rather than the type of the incision.

Kirali et al. [6] conducted a small study comparing reversed-J inferior sternotomy (n = 10) and full sternotomy (n = 7) in awake patients using a high thoracic epidural anaesthetic protocol. While both groups demonstrated a restrictive reduction in lung function compared with preoperative values, there was no statistical difference in postoperative values.

Ohkado et al. [7] investigated respiratory function in 30 patients divided into three groups (MIDCAB, OPCAB and conventional CABG). A significant difference was seen in vital capacity 2 weeks postoperatively in the OPCAB (94.2 ± 4.6 vs 77.3 ± 6.3%) and CABG (96.9 ± 3.7 vs 78.2 ± 4.2%) groups when compared with preoperative values, but not in the MIDCAB group (94.0 ± 7.5 vs 89.1 ± 6.2). The authors suggest that the MIDCAB approach is related to better respiratory recovery postoperatively.

Lichtenberg et al. [8] compared two groups of 15 men undergoing MIDCAB and conventional CABG with CPB. There were no significant differences preoperatively or on postoperative Day 1. However, by postoperative Days 3 and 5, the conventional CABG group had worse vital capacity and %FEV₁. Higher postoperative pain was reported in the MIDCAB group on Day 3. They concluded that respiratory function recovered more rapidly in the MIDCAB group despite the higher levels of pain initially (on forced expiration, but not at rest), and therefore, patients with impaired respiratory function are recommended for MIDCAB where possible.

Guizilini et al. [9] compared lung-function tests following standard sternotomy or ministernotomy in 18 patients. FEV and FVC were higher in the ministernotomy group postoperatively when compared with preoperative levels. Furthermore, less shunt, less postoperative pain and a shorter hospital stay were seen in the ministernotomy group. The study used the same practitioner to supervise the spirometry, and the same LITA technique and general anaesthetic protocol between both groups.

Grossi et al. [10] compared lung-function tests and pain in patients undergoing CABG by port access or standard sternotomy in a 6-month period. The port-access group showed better recovery of lung function up to postoperative Day 3, but this difference lost statistical significance by the end of the 6-week follow-up period. The study also looked at a comprehensive range of outcomes including markers of postoperative stress, but the design allowed the surgeon to choose their approach, leaving a potential for bias.

There is significant heterogeneity in the studies reported above, including length of follow-up and timing of lung-function testing. Importantly, these studies can be split into ministernotomy and thoracotomy. Randomized evidence suggests no clear benefit to using a ministernotomy approach in terms of lung function. The cohort studies for ministernotomy are split: one suggests an improvement [10], while the other concurs with the randomized study [9]. Randomized evidence for the use of thoracotomy is divided, with one study reporting no benefit attributable to reduced surgical trauma but that it appears safe [5], and another suggesting a worsened lung function [3]. The cohort studies for thoracotomy suggest a benefit to respiratory function, particularly in the early days of recovery.

CLINICAL BOTTOM LINE

There is no randomized evidence to suggest that minimally invasive surgical approaches confer a benefit to postoperative respiratory function over conventional median sternotomy. However, small non-randomized studies support the benefit of minimally invasive surgery.

Conflict of interest: none declared.