Minimally invasive left internal mammary artery harvesting techniques during the learning curve are safe and achieve similar results as conventional LIMA harvesting techniques.

<bold>Background: </bold>Internal thoracic arteries (ITAs) are considered to be the standard conduits used for coronary revascularization. Recently minimally invasive procedures are performed to harvest ITAs. The aim of this retrospective cohort study is to observe the effect and safety of less invasive LIMA harvesting approaches in the learning curve compared to conventional harvesting.<bold>Methods: </bold>We retrospectively analyzed the data of 138 patients divided into three different groups based on the LIMA harvesting techniques: conventional sternotomy LIMA harvesting, CSLH (n: 64), minimally invasive direct LIMA harvesting, MIDLH (n: 42), and robotic-assisted LIMA harvesting, RALH (n: 32). The same 138 patients were also divided into sternotomy (n: 64), and non-sternotomy (n: 74) groups keeping both MIDLH and RALH in the non-sternotomy category. Parameters associated with LIMA's quality and some other perioperative parameters such as harvesting time, LIMA damage, perioperative myocardial infarction, ventilation time, 24 h drainage, ICU stay, hospital mortality, computed tomographic angiography (CTA) LIMA patency on discharge, and after one year were recorded.<bold>Results: </bold>The mean LIMA harvesting time was 36.9 ± 14.3, 74.4 ± 24.2, and 164.7 ± 51.9 min for CSLH, MIDLH, and RALH groups respectively (p < 0.001). One patient 1/32 (3.1%) in the RALH group had LIMA damage while the other two groups had none. One-month LIMA CTA patency was 56/57 (98.2%), 34/36 (94.4%), and 27/27 (100%) (p = 0.339), while 1 year CTA patency was 47/51 (92.1%), 30/33 (90.9%), and 24/25 (96%) for CSLH, MIDLH, and RALH groups respectively (p = 0.754). In the case of sternotomy vs non-sternotomy, the LIMA harvesting time was 36.9 ± 14.3 and 113.6 ± 59.3 min (p < 0.001). CTA patency on discharge was 56/57 (98.2%) and 61/63 (96.8%) (p = 0.619), while 1 year CTA patency was 47/51 (92.1%) and 54/58 (93.1%) (p = 0.850) for sternotomy vs non-sternotomy groups.<bold>Conclusion: </bold>Minimally invasive left internal mammary artery harvesting techniques during the learning curve are safe and have no negative impact on the quality of LIMA. Perioperative outcomes are comparable to conventional procedures except for prolonged harvesting time. RALH is the least invasive and most time-consuming procedure during the learning curve. These procedures are safe and can be performed for selected patients even during the learning curve. [ABSTRACT FROM AUTHOR]