Objective To compare the short-term efficiency of two different endovascular repairs for type B aortic dissection involving the left subclavian artery.MethodsFrom February 2013 to March 2016, a cohort of 43 patients with TBADs involving the LSA underwent thoracic endovascular aortic repair (TEVAR) in two departments, consisting of 22 (Group A) with chimney grafts (CGs) and 21 (Group B) with single-branched stent graft (SBSG). Results of the two groups in perioperative and follow-up period (≥ 3 months) were comparatively analyzed, especially on aortic remodeling.ResultsEndoluminal repair of the two groups was successfully carried out. The median follow-up period was 19 months (range, 3–43 months) in Group A and 12 months in Group B (range, 6–32 months). During the TEVAR, one CG compression occurred in Group A and one type I endoleak in Group B. During follow-up, four complications occurred in Group A (two CGs occlusion, one type I endoleak and one death from dissecting aneurysm rupture), compared with two occurred in Group B (one sidearm graft twist and one death from myocardial infarction). Complete thrombosis of the false lumen (FL) in thoracic aorta was revealed in 83.3% (15/18) cases in Group A and 89.5% (17/19) in Group B. Partial thrombosis of the FL was revealed in 16.7% (3/18) cases in Group A and 10.5% (2/19) in Group B. In the abdominal aorta, complete thrombosis of the FL was noted in 23.1% (3/13) cases in Group A and 36.4% (4/11) in Group B. Partial thrombosis of the FL was revealed in 76.9% (10/13) cases in Group A and 63.6% (7/11) in Group B. Significant true lumen re-expansion and false lumen regression were observed in different levels of the descending aorta by computed tomography angiography (CTA) in both Groups A and B (P < 0.05). No significant diametric changes of abdominal aorta were found during follow-up in both groups.ConclusionsFor patients with TBADs involving the LSA, the chimney technique and the SBSG revealed comparable results. Further evaluation of more patients with longer follow-up is needed to substantiate these results.