Discussion
IMH is part of the spectrum of acute aortic syndromes (5–25%) and is more common in the descending thoracic aorta (type B, 60–70%) than in the ascending aorta [1-3] . IMH without malperfusion syndrome is always regarded as a benign disease, and the “wait-and-watch strategy” (initial medical treatment plus necessary thoracic endovascular aortic repair treatment) is the first recommendation for these patients [1-3][17]. Many potential risk factors for IMHB have also been summarized, including a maximum aorta diameter larger than 45 mm, increased pleural effusion, hematoma thickening larger than 10 mm, ULP development, and elevated CRP levels [1]. Overall, patients with IMHB have a more favorable long-term prognosis than patients with aortic dissection and an in-hospital mortality risk lower than 10% during the acute phase[1]. The mortality rate after TEVAR during the acute phase in our study was 18% in the non-DM group (Table 2 ), similar to the report by Schoenhoff et al. (the mortality rates of TEVAR within and after the first two weeks were 25% and 29%, respectively) and was higher than that of type B aortic dissection patients who underwent TEVAR treatment (0% to 18%, and median 6%) [5] . Falconi et al. [18] reported that of 27 type B IMHs in patients managed conservatively and followed for a mean of 33 months, 47% underwent regression, 14% remained stable, and 39% progressed to aortic dissection or enlargement. Motoyoshi et al. [18]reported that of 26 patients managed medically, 6 patients (23%) had spontaneous regression, and 7 patients (27%) required a surgical procedure. Although the non-DM group in our study had a higher reintervention rate than the DM group during the follow-up (11% vs 2%,P =0.044), the reinvention rate was obviously lower than that in previous studies (60% [6] , 39% [18] , 27%[19] ).
Newly diagnosed type 2 DM patients with IMHBs probably benefit from antidiabetic treatments, and tight glycemic control may influence IMHB evolution. The potentially protective value of DM has been well described [8-11], and the possible explanations include the increasing matrix of the aortic wall (suppression of plasmin and decreased levels/activity of MMP-9) and decreasing aortic mural macrophage infiltration and neovascularization [16] . MMP-9 is involved in tissue degradation and remodeling in aortic dissection and is significantly increased in aortic dissection patients, and a higher level of MMP-9 can weaken the aortic media by degrading multiple extracellular components; DM patients have a 2-fold decreased level of MMP-9, which may restrict degradation of the aortic wall[16]. Unlike the effect of hyperglycemia on immune cell activity in type 2 DM, the impact of insulin deficiency in type 2 DM on macrophage activity has not been widely studied. Tessaro et al. demonstrated that the administration of exogenous insulin in diabetes may enhance the immune activity of macrophages [20], and insulin treatment may diminish this protective effect of hyperglycemia that prevents aortic aneurysm development (under laboratory conditions)[21]. For these reasons, it seems that the administration of insulin is probably a risk factor for disease progression in type 2 DM patients with IMHBs. In our study, after receiving tight glycemic control recommended by guidelines [7], patients in the DM group had an MMP-9 level that was dramatically increased (reached the highest value) during the acute phase, and the MMP-9 level reached the highest level earlier in the DM group than in the non-DM group. However, the MMP-9 levels in the DM group were lower than those in the non-DM group at each timepoint (P< 0. 001, Figure 2 ), and patients in the non-DM group had a 2-fold higher level of MMP-9 after day 90 than patients in the DM group, which likely indicated that the potentially protective effect of hyperglycemia did not decrease after the administration of insulin, which could likely explain the significantly lower aorta-related mortality in the DM group during the follow-up period. Until now, the mechanism of the protective effect of hyperglycemia and the administration of insulin and other antidiabetic treatments in aortic diseases has remained unclear and requires further study.
For this higher mortality rate in the non-DM group than in the DM group after TEVAR treatment, one possible explanation was the larger aorta diameters and greater hematoma thickness in the non-DM group. Bischoff et al. [22] indicated that patients with larger ascending aorta diameters had a worse clinical outcome than those with smaller aortas. Ye et al. also recommended that patients with aorta diameters larger than 45 mm and hematoma thicknesses greater than 10 mm should be included in the “complicated” group, and more aggressive TEVAR treatment was required in these patients to prevent fatal IMHB evolution[23] . In addition, the geometry of the aorta can be influenced by TEVAR, which can result in a poor outcome, especially in those with ascending aortas larger than 40 mm (who are more likely to suffer from retrograde type A aortic dissection after TEVAR)[24] . In the non-DM group, one patient died of retrograde type A aortic dissection after TEVAR during the acute phase. Both patients who died after TEVAR in the non-DM group had ascending aorta diameters larger than 40 mm (Supplement 1 ). Thus, an ascending aorta diameter larger than 40 mm is another potential evolutionary predictor of IMHB.
The other possible explanation for this higher mortality rate in the non-DM group is the improper timing of TEVAR and the stent graft size. The current timing of TEVAR for uncomplicated IMHB is in line with that for uncomplicated type B aortic dissection (patients with IMH expansion despite medical therapy and ULP development) [1][4] . However, the disease progression of IMHB is highly unpredictable, and the current guidelines are based on good outcomes of TEVAR for type B aortic dissection[25][26] and on some small sample studies for IMHBs[1][4][27][28] . In our study, all 12 patients with uncontrollable symptoms (especially refractory pain) underwent TEVAR. The two patients who finally died of aorta-related complications still complained of refractory pain after TEVAR (Supplement 1 ). Juvonen et al. [29] reported that patients with uncharacteristic or atypical pain had a higher risk of rupture over time than those without pain. Medical management of such patients is unwarranted unless their life expectancy and quality of life are markedly impaired [30] . A more aggressive TEVAR procedure (while the symptoms are still controllable) would likely have better outcomes (acute phase mortality: 0%, Bischoff et al. [22] , 0%, Ye et al. [23] ), and this more aggressive strategy is likely to prevent the progression of IMHB and result in a better prognosis than medical treatment plus necessary TEVAR (for patients with uncontrollable symptoms and fatal evolution). Further studies are necessary to evaluate whether patients with uncomplicated IMHBs without diabetes mellitus could benefit from more aggressive TEVAR treatment while the symptoms are still controllable. In addition, in our study, the diameter of the stent graft was oversized by less than 10%, which was smaller than the diameters in the study by Schoenhoff et al. (10% to 15%). [5]. However, both in our study and in the study by Schoenhoff et al. [5] , the mortality rates were higher than those after TEVAR for type B aortic dissection[4][31] . For aortic dissection, there is a general agreement in the current guidelines that the oversizing factor should not exceed 5% [31] . Is it necessary to utilize less oversizing (not exceeding 5% [31] ) for stent grafts similar to the recommendations for aortic dissection in patients with IMHB?
In addition, ULP development during the acute phase and CRP levels were associated with an elevated risk for aorta-related mortality. Moral et al. [32] reported that 10% of patients with uncomplicated IMHBs suffered from ULP development during the acute phase; among these patients, the disease progression rate was as high as 91%, and 36% of them died of aortic-related complications. Kitai et al. reported that 71% of patients with IMHBs had ULPs during the acute phase, of which 76% showed progression (enlargement or progression to an aortic aneurysm) and required further treatment [33] . The potential risk factors for ULP development are large aorta diameters and hematoma thicknesses, which are also evolutionary predictors of IMHB[30]. In addition, a higher CRP level may be indirect evidence that could represent the degree of inflammation and pathophysiological changes in the aortic wall under such unstable hemodynamic conditions; a sustained, high CRP level has significant prognostic value in IMHB patients [34].
There are several limitations to this study. First, in this study, the majority of patients received only two CTA examinations during the acute phase (once upon admission and once before discharge), and the diagnosis of ULPs could have been influenced by the preciseness of the CTA images. Closer monitoring for the development of ULPs is necessary in further studies. Second, in future studies, we should enroll more patients with different medical treatment strategies, which may provide more meaningful insight into the influence of these strategies on IMHBs.