Mini‐sternotomy vs right anterior thoracotomy for aortic valve replacement

While minimally invasive techniques for aortic valve replacement (AVR) have been shown to be safe, limited data exist comparing the varying approaches. This study aimed to compare the outcomes between two minimally invasive approaches for AVR: mini‐sternotomy (MS) and right anterior thoracotomy (RAT).


| BACKGROUND
With the increasing prevalence of aortic stenosis in the elderly population, 1,2 the need for less traumatic approaches for treatment are becoming increasingly important. Minimally invasive aortic valve replacement (MIAVR) has been shown to achieve similar mortality rates to conventional aortic valve replacement, albeit with more technical demand. 3 MIAVR uses a smaller incision and avoids complete division of the sternum, conferring several benefits: lower ventilation time, 4-6 pain scores, 5,6 intensive care unit (ICU) stay, 5,7 Records idenƟfied through database searching (n = 1,013) and hospital stay. 7,8 There is also evidence to suggest a lower transfusion requirement during surgery, 5 and a reduced volume of blood lost from chest drainage. 4,6 Mini-sternotomy (MS) and right anterior thoracotomy (RAT) are widely reported as minimally invasive approaches for AVR. MS is performed via a 6-to 10-cm midline skin incision with J sternotomy at the third to fourth intercostal space, whereas RAT is performed via a 5-to 7-cm incision in the right second intercostal space. 9 RAT sacrifices the right internal thoracic artery and reduces the operative field more so than MS. 3 Current US and European guidelines do not give preference to either procedure 10

| Study selection
Three review authors (LC, HH, and ID) independently assessed titles and abstracts of all the papers found in the electronic search. Potential studies were marked as "retrieve" and any differences between reviewers were discussed. The full text publications were retrieved and assessed, and were included if they 1 reported ≥1 of the predetermined outcomes of the review for both MS and RAT 2 ; were human studies published in English. Abstracts were excluded as there was insufficient methodological reporting to allow for risk of bias assessment.

| Data extraction
A data collection form was used for study characteristics and outcome data. The following characteristics were extracted.  This formula provides a near unbiased estimation of SD for normally distributed data and a 5% relative error for skewed data. Sensitivity analysis was then performed by removing these studies to determine if they had influenced the overall direction of the effect estimate.

| Sensitivity and heterogeneity
Sensitivity analysis was performed by removing studies that had a particularly high risk of confounding. Substantial heterogeneity was considered as a P-value of the χ 2 statistic of <0.10 or an I 2 statistic of greater than 50%.

| Assessment of in-study bias
Risk of bias was assessed independently by two reviewers (HH and PR) using the Newcastle-Ottawa scale (NOS). 15 The NOS is based on three components of study methodology: participant selection, comparability of cohorts, and measurement of outcomes. Funnel plot analysis assessed for publication bias. Another cluster of three studies also came from the same center using patients in similar time periods. [18][19][20] The study by Gilmanov 19 appeared to cross-over in its patients with the earlier Miceli study 18 and was therefore excluded.

| Included studies
Nine observational studies were identified that met the inclusion criteria; one prospective 21 and eight retrospective cohort studies. 16

| Risk of bias in included studies
Risk of bias was assessed using the NOS (Figure 2), which has a maximum score of 9. Two studies scored 8, 21 20 The "risk of bias summary" table provides a breakdown of the scoring ( Figure 1B). Funnel plot analysis found little evidence of publication bias, while Egger's test found no small-study effects (P = .127) ( Figure 1C).

| Intraoperative parameters
The rate of conversion to sternotomy was assessed in four studies. 18

Re-operaƟon for bleeding
(C) It is widely agreed that AVR via RAT is a more technically demanding procedure, with a more limited operating field, and this has been reflected in our analysis: surgeons were more likely to convert to a full sternotomy in RAT patients compared to MS patients, undoubtedly to improve the access in demanding cases where the surgeon deemed minimal access as a hindrance operative progress.
Our analysis found increased AoX times in RAT patients, reflecting the challenge this access poses to timely performance of the various steps of the AVR procedure. Despite this, surgical experience is likely to vary among the studies we have reported.
As with most minimally invasive procedures, a recognized learning curve exists. Studies have shown that and cautious management of the learning curve in surgical procedures can lead to improved outcomes. 27,28 Variability in experience and proficiency among surgeons affects the number of cases required to overcome the learning curve. 29 While most reports in our analysis emerge from specialized centers, the overall recommendation is for the operative approach to be selected according to surgeon's technical expertise, which in turn may affect the observed outcomes.

| Strengths and limitations
The analysis offered by this study provides an improved comparison for two viable methods of approaching the aortic valve with increasing acceptability. The overall quality of evidence, however, is