This study was performed to assess postoperative aortic remodelling (AR) after total arch replacement (TAR) for acute type A aortic dissection (AAD) with a frozen elephant trunk (FET) or conventional elephant trunk (cET). Furthermore, the shape of the residual true lumen (TL) was analyzed based on elliptical Fourier analysis (EFA) and evaluated as a predictor of AR. This study involved patients who underwent TAR with a cET or FET for AAD from December 2006 to January 2023 at five institutions. AR was assessed at the levels of the 4th thoracic vertebra (Th4), Th7, Th10, and above the coeliac trunk. The shape of the residual TL at all four levels was analyzed based on EFA to calculate shape patterns as principal component (PC) values. Inverse probability of treatment weighting (IPTW) was performed for adjustment between the groups. In total, 180 patients (88 with cET and 92 with FET) were enrolled. The complete AR rate, defined as false lumen remodelling throughout the entire descending thoracic aorta, was significantly higher in the FET than cET group (63.4% vs 32.0%, P = 0.0013). The IPTW-adjusted Fine-Gray regression model revealed that the mean PC2 (hazard ratio, 0.22; P < 0.001) and PC3 (hazard ratio, 0.24; P = 0.009) of the four levels were independent predictors of complete AR. In AAD repair, the AR rate was significantly higher with use of the FET than cET. The shape patterns of the residual TL can be an important reference for predicting postoperative AR.