Clinical outcomes and follow-up results after matching
In the DM group, ten patients received TEVAR treatment, including two patients with ulcer-like projection development, seven patients with type B aortic dissection development, and one patient with aortic pseudoaneurysm. Two patients developed a type A aortic dissection in the DM group and received open surgery; one patient died after the surgery (Table 2 and Figure 3 ). In the no DM group, nine patients with intramural hematoma evolution and unstable syndromes underwent TEVAR (including one case of ulcer-like projection development, seven cases of type B aortic dissection and one case of aortic pseudoaneurysm, Figure 3 ). There were no statistically significant different in the procedural details between the two groups, and the stent was successfully deployed in all nineteen patients (Table 2 ). The aorta-related mortality during the first hospital stay was similar among two groups (9.1% vs 6.1%,P =1.000, Table 2 ). The indications and outcomes of surgery/TEVAR are summarized in Figure 3. There were no statistically significant differences in the procedural details between the two groups (Table 2 ). All 61 patients survived the in-hospital had follow-up after discharge. In the no DM group, two non-aorta-related death cases involved patients who died of renal failure at 40 months and 39 months and the only non-aorta-related death case in the DM group involved a patient who died of lung cancer at 37 months. The Fine-Gray model for the competing risk analysis revealed a significantly higher aorta-related mortality during the follow-up period in the no DM groups than in the DM group (36.4%; 95% CI, 11.6%-82.3%, P= 0.0294). The cumulative incidence curves for the two groups were significantly different for aorta-related death (P =0.0294) and not significantly different for non-aorta-related death (P =0.567) (Figure 4A-C ). In the no DM group, more aorta-related deaths occurred during the 3 to 6 months after the onset of intramural hematoma (P =0.011) (Figure 4D ). Patients without DM experienced significantly more aorta-related adverse events (51.6% vs 13.3%, P =0.001) and reinterventions than patients in the DM group (29.0% vs 6.7%, P =0.023) (Table 2 ). In the no DM group, the development of aortic dissection was the most common reason for TEVAR/surgery reinterventions (n=10), and the majority of aorta-related death cases (six patients) occurred during the 3 to 6 months after the onset of intramural hematoma (Figure 4D ). Five of these six patients (Patients 5, 6, 8, 9 and 10, Supplement 2 ) in the no DM group suffered from chest/back pain after TEVAR during the subacute phase (14-90 days)(P <0.014) (Table 2 ) and died of retrograde type A aortic dissection or ascending aortic pseudoaneurysm rupture(Supplement 2 ). During follow-up, eight in ten deaths were caused by rupture of a retrograde type A aortic dissection (n=5) and ascending aortic pseudoaneurysm (n=3) after receiving TEVAR (Patients D -G , Supplement 3 ).