Topologic distributions of vasa vasorum and lymphatic vasa vasorum in the aortic adventitia – Implications for the prevalence of aortic diseases

Atherosclerosis 247 (2016) 127e134

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Topologic distributions of vasa vasorum and lymphatic vasa vasorum in the aortic adventitia e Implications for the prevalence of aortic diseases Masaki Sano a, b, Naoki Unno a, b, *, Takeshi Sasaki c, Satoshi Baba d, Ryota Sugisawa a, b, Hiroki Tanaka a, b, Kazunori Inuzuka a, b, Naoto Yamamoto a, b, Kohji Sato c, Hiroyuki Konno b a

Division of Vascular Surgery, Hamamatsu University School of Medicine, Japan Second Department of Surgery, Hamamatsu University School of Medicine, Japan c Department of Anatomy and Neuroscience, Hamamatsu University School of Medicine, Japan d Department of Diagnostic Pathology, Hamamatsu University School of Medicine, Japan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 September 2015 Received in revised form 27 January 2016 Accepted 3 February 2016 Available online 12 February 2016

Background and aims: Vasa vasorum (VV) and lymphatic vasa vasorum (LVV) form their own networks in the adventitia. VV supply the aorta with nutrition and oxygen; however, the distribution and role of LVV remains to be determined. The purpose of this study was to investigate differences in the distribution of VV and LVV along the aorta. Methods: Aortic samples were obtained from 22 autopsy cases without medical history of aortic diseases. Aortic segments were classified as arch (Ar), descending thoracic (De), suprarenal abdominal (S-Ab), and infrarenal abdominal (I-Ab). Adventitial VV and LVV were identified immunohistochemically. Results: VV were most dense in the arch aorta, becoming less dense along the aorta in more distal segments, with the lowest density occurring in the infrarenal abdominal aorta. There was a significant correlation between the numbers of VV and medial thickness in the total aortic segments (r ¼ 0.518, p < 0.01). In contrast, there was no significant correlation between the number of LVV and medial thickness in any aortic segments. However, there was a significant correlation between the number of LVV and intimal thickness in I-Ab (r ¼ 0.425, p < 0.05). Conclusions: The distributions of adventitial VV and LVV were characteristic along the aortic segments. Differences in the distributions may imply the prevalence of aortic diseases such as dissection, abdominal aortic aneurysm, and atherosclerotic occlusive disease in each aortic segment. © 2016 Elsevier Ireland Ltd. All rights reserved.

Keywords: Abdominal aorta Lymphatic vasa vasorum Distribution Intimal thickness Dissection Aneurysm Occlusive disease

1. Introduction Aortic diseases are diverse, are associated with multiple biological systems in the aortic wall, and interface with the blood. The prevalence of aortic diseases varies along the aorta, including the ascending aorta, aortic arch, descending aorta, thoracoabdominal

* Corresponding author. Division of Vascular Surgery, Hamamatsu University School of Medicine, 1-20-1 Handayama, Higashi-ku, Hamamatsu 431-3192, Japan. E-mail addresses: [email protected] (M. Sano), [email protected] (N. Unno), [email protected] (T. Sasaki), [email protected] (S. Baba), r. [email protected] (R. Sugisawa), [email protected] (H. Tanaka), inu@ hama-med.ac.jp (K. Inuzuka), [email protected] (N. Yamamoto), ksato@ hama-med.ac.jp (K. Sato), [email protected] (H. Konno). http://dx.doi.org/10.1016/j.atherosclerosis.2016.02.007 0021-9150/© 2016 Elsevier Ireland Ltd. All rights reserved.

aorta, and abdominal aorta. For instance, atherosclerotic aneurysms are at least three times more frequent in the abdominal aorta than in the descending thoracic aorta [1]. In addition, more than 94% of abdominal aortic aneurysms (AAA) occur in the infrarenal region of the aorta in humans. Moreover, atherosclerotic aortic stenosis is much more frequent in the abdominal aorta than in the descending thoracic segment of the aorta [2]. In contrast, 65% of cases of aortic dissection occur in the ascending aorta, with 20% in the descending aorta, 10% in the aortic arch, and only 5% in the abdominal aorta [3]. This tropism of aortic diseases may be attributed to differences in the biological architecture of the aorta, particularly the perfusion system of the aortic wall. The nutritional supply of the internal part of the aorta is provided by perfusion from the aortic lumen, and that of the external part is provided by the adventitial vasa vasorum

128

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

was as follows: hypertension (medication for hypertension or a systolic blood pressure > 140 mmHg and/or a diastolic blood pressure > 90 mmHg), hypercholesterolemia (medication for hypercholesterolemia or total serum cholesterol concentration > 220 mg/dL), hypertriglyceridemia (medication for hypertriglyceridemia or total serum triglyceride > 150 mg/dL), diabetes (present or past medication for diabetes), and smoking (present or past smoking history). The distal arch (Ar), and descending thoracic (De), suprarenal abdominal (S-Ab), and infrarenal abdominal (I-Ab) segments, as well as mesenteric lymph nodes were resected (Fig. 1A). Samples were immersed in 10% neutral buffered formalin for 24 h and embedded in paraffin.

(VV) [4]. Almost a half century ago, the distribution of VV in the abdominal aorta was identified to be much more scarce than that in the thoracic aorta [5]. Recent studies describing the difference of the inherent VV distribution between the thoracic and abdominal aorta may explain differences in the pathogenesis of thoracic and abdominal aortic aneurysms [6e9]. Thus, the distribution of VV along the aorta may be associated with the diversity of aneurysmal diseases in the aorta. Another micro-vessel structure in the human arterial adventitia, the lymphatic vasa vasorum (LVV) has long been neglected by researchers owing to difficulties with identification within a specimen. However, the development of immunohistological techniques and the availability of antibodies against lymphatic vessels have helped researchers distinguish luminal structures between VV and LVV. Since then, adventitial LVV have very recently been found in the human carotid artery, iliac artery, and aorta [10,11]. Because the lymphatic system plays critical roles in transporting lipids, cytokines, and inflammatory cells in the arterial wall, differences in the LVV distribution along the aorta may also be associated with the topological diversity of aortic diseases. In this study, we investigated human aortae using autopsy samples and identified differences in the distributions of VV and LVV along the aorta.

2.2. Histological and immunohistochemical staining Thin sections were cut, and Elastica van Gieson (EVG) and immunohistochemical staining for mouse monoclonal antibody against von Willebrand Factor (vWF) (1:200, DakoCytomation, Grostrup, Denmark) and mouse monoclonal antibody against podoplanin (1:200, DakoCytomation) was performed [12,13]. Slides were stained with the appropriate secondary antibody and counterstained with hematoxylin. Double immunofluorescence staining for mouse monoclonal antibody against podoplanin (1:200, DakoCytomation) and rabbit polyclonal antibody against human lymphatic vessel endothelial hyaluronan receptor (LYVE)-1 (1:100, ReliaTech, Wolfenbüttel, Germany) was performed to identify lymphatic endothelium [12,13,14]. Slides were stained with the appropriate secondary antibody and DAPI.

2. Methods 2.1. Sample collection We collected samples from autopsy cases at the Department of Diagnostic Pathology, Hamamatsu University Hospital, between October 2011 and August 2014. All procedures were approved by the Ethics Committee of Clinical Research of the Hamamatsu University School of Medicine, and informed consent was obtained from the donor's family. We sampled the dorsal side aortic walls (almost half of the circumference), because visceral arteries, which are considered to be origins of VV, are located at the ventral or lateral side of the aorta. Autopsy cases with collagen disease, Marfan syndrome, or aortic degenerative diseases (aneurysm, dissection, and atherosclerotic stenosis) and those under 18 years of age, were excluded. Moreover, aortic samples with calcific plaque were also excluded. Twenty-two autopsy cases were studied; each patient's demographics, cause of death, and co-morbid conditions (hypertension, hypercholesterolemia, hypertriglyceridemia, diabetes, smoking, ischemic heart disease, and cerebrovascular disorder) are shown in Tables 1 and 2. The definition of each condition

2.3. Morphometric analysis Based on the analysis of four sections at 500 mm intervals, the number of VV and LVV in 50 view fields was counted under a light microscope (200). The adventitial area was defined as the area between the external elastic lamina and either the edge of the periaortic adipose tissue or 300 mm from the external elastic lamina. Intimal thickness was defined as the distance from the lumen to the internal elastic lamina, and medial thickness was defined as the distance from the internal elastic lamina to the external elastic lamina; distances were measured using a computerized image analysis system (Lumina Vision version 3.0, Mitani Corp. Tokyo, Japan).

Table 1 Clinical characteristics of autopsy cases. Autopsy cases (n ¼ 22) Sex (M/F) Age BMI Hypertension Hypercholesterolemia Hypertriglyceridemia Diabetes Smoking Ischemic heart disease Cerebrovascular disorder

þ  þ  þ  þ  þ  þ  þ 

10/12 64.6 ± 3.1 20.8 ± 0.7 12 10 9 13 7 15 5 17 9 13 2 20 5 17

Table 2 Autopsy cases. BMI Hypertension Hypercholesterolemia (>220 mg/dl)

1 2 3 4 5 6 7 8 9 10 11

82 67 62 83 58 73 31 61 69 51 46

M M F M F M F F M M F

19.3 19.4 17.1 20.0 21.0 18.0 20.2 26.1 15.3 26.0 22.9

12

79

F

13 14

59 68

F F

15 16 17 18 19 20 21 22

88 56 69 67 79 63 86 44

F M F M M M F F

MRSA infection Pancreatic carcinoma Multiple myeloma Myocardial infarction Bladder carcinoma Interstitial pneumonia Malignant glioma Mesenteric liposarcoma Pancreatic carcinoma Pancreatic carcinoma Pulmonary thromboembolism Gastrointestinal hemorrhage Hepatic angiosarcoma Gastrointestinal stromal tumor Aortic valve stenosis Pleural empyema Interstitial pneumonia Multiple myeloma Interstitial pneumonia Liver cirrhosis Lung carcinoma Suppurative cholangitis

þ   þ    þ þ þ 

 þ  þ   þ   þ 

  þ    þ þ  þ 

þ þ  þ       

þ þ  þ  þ  þ  þ 

   þ       

     þ   þ  

The number of The number of VV/50 view fields LVV/50 view fields Ar De S- I- Ar De S- IAb Ab Ab Ab 335 378 411 307 34 5 17 51 487 438 417 252 16 16 31 61 348 196 306 206 23 18 26 31 419 324 281 325 6 4 9 13 393 346 221 229 3 10 13 35 347 348 311 308 23 15 68 27 427 411 441 298 44 5 38 37 289 398 255 288 30 12 13 21 391 384 309 282 31 13 51 34 282 297 323 365 22 5 12 52 406 271 264 318 19 17 22 23

24.0 þ











þ

300 378 315 285 18 21 100

55

20.3  24.2 þ

 þ

 

 

 

 

 

343 256 389 338 19 7 326 291 259 216 17 10

6 28

þ  þ þ  þ þ 

  þ þ þ   þ

 þ  þ    þ

   þ    þ

 þ  þ  þ  

   þ    

    þ þ  

409 353 508 321 257 571 451 389

16.9 20.7 22.6 21.2 20.1 30.5 20.9 30.1

Hypertriglyceridemia (>150 mg/dl)

Diabetes Smoking Ischemic heart disease

Cerebrovascular disorder

290 341 443 389 325 292 519 317

350 355 455 370 380 369 380 316

297 277 354 279 286 258 299 284

9 39 38 39 25 29 66 12

17 11 6 17 13 15 10 14

7 8

6 3 4 26 49 60 36 29 24 15 48 118 43 15 16 21

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

Case Age Sex Cause of death

129

130

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

Fig. 1. Distributions of adventitial vasa vasorum (VV) and lymphatic vasa vasorum (LVV) in the aorta. A: Schematic illustration of aorta. Ar: distal arch aorta, De: descending aorta, SAb: suprarenal abdominal aorta, I-Ab: infrarenal abdominal aorta, SMA: superior mesenteric artery. B: Elastica van Gieson (EVG) staining of normal abdominal aortic wall. C: Higher magnification of the outlined area in B. D: Adventitial vasa vasorum (VV) positively stained with anti von Willebrand Factor (vWF) antibody (arrows). E: adventitial lymphatic vasa vasorum (LVV) positively stained with anti-podoplanin antibody (arrows). F: Double immunofluorescence staining of mesenteric lymph nodes, G: Double immunofluorescence staining of aortic adventitia (red: podoplanin, green: LYVE-1, blue: DAPI). Scale bars indicate 200 mm (B), 100 mm (C-E), and 10 mm (F, G). H: The average of the number of VV in each aortic segment in 50 view fields under high magnification (200). Means ± standard deviations are shown. I: The average of the number of LVV in each aortic segment in 50 view fields at high magnification (200). Means ± standard deviations are shown. J: The average of the ratio of LVV to VV. Means ± standard deviations are shown. *p < 0.05, **p < 0.01.

2.4. Statistical analysis The numbers of VV and LVV, intimal thickness, and medial thickness are expressed as means ± standard deviation. Differences in these measurements between each aortic segment were analyzed using the Kruskal-Wallis test. Statistical significance was defined as a p -value < 0.05. The relationships between the presence of co-morbid conditions (hypertension, hypercholesterolemia, hypertriglyceridemia, diabetes, smoking, ischemic heart disease, cerebrovascular disorder) and the number of VV or LVV in each aortic segment were analyzed using logistic regression analysis. Scatter plots were used to identify correlations between intimal or

medial thickness and the number of VV or LVV in each segment. These correlations were evaluated using correlation coefficient analysis. A strong correlation was defined as a correlation coefficient (r) > 0.4. All statistical analyses were performed using SPSS 17.0 for Windows (SPSS Inc., IL, USA). 3. Results 3.1. Immunohistochemical staining of vasa vasorum and lymphatic vasa vasorum in aortic adventitia EVG staining revealed a wavy structure of elastin fibers in aortic

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

media and adventitial vessels (Fig. 1B, C). Immunohistochemical staining revealed adventitial VV (Fig. 1D) and LVV (Fig. 1E). The nutritional supply of the external part of the aortic media in Ar and De is provided by medial VV [4]. VV were often observed in aortic media of Ar and De, but were seldom observed in that of S-Ab and IAb. LVV were often observed near VV in adventitia, but were seldom observed in intima and media. Podoplanin has also been reported to be found in normal cells other than lymphatic endothelial cells [15]. Immunohistochemical staining for the other lymphatic-specific marker LYVE-1 was performed to indicate lymphatic endothelial cells. Double immunofluorescence staining showed that normal lymphatics in mesenteric lymph nodes expressed podoplanin and LYVE-1 (Fig. 1F). Adventitial LVV also expressed podoplanin and LYVE-1 (Fig. 1G). 3.2. Number of adventitial VV and LVV in each aortic segment There was no significant difference between the presence of each co-morbid condition and the number of VV or LVV in each aortic segment. The distribution of VV was most dense in Ar (p < 0.05) and decreased as the aortic segments became more distal, with the lowest amount occurring in I-Ab (Fig. 1H). In contrast, the distribution of LVV was most dense in I-Ab among the four segments of the aorta (p < 0.01; Fig. 1I). Moreover, the LVV/VV ratio showed that the distribution of LVV in comparison with VV was most prominent in I-Ab (p < 0.01; Fig. 1J). 3.3. Intimal thickness and medial thickness in each aortic segment Intimal thickness in each aortic segment was greatest in I-Ab and least in Ar (Fig 2A, B), whereas medial thickness was greatest in Ar and decreased along the aorta, with the lowest values occurring in I-Ab (Fig. 2A, C). 3.4. Correlations between the distribution of VV or LVV and aortic intimal thickness or medial thickness Scatter plots showed that there were no correlations between

131

the number of VV and intimal thickness in any aortic segment (Fig. 3A-E). On the other hand, there were significant correlations between the numbers of VV and medial thickness in all aortic segments except De (Fig. 3F-J). There was a significant correlation between the number of VV and medial thickness among the total aortic samples (Fig. 3J). With regard to the distribution of LVV, there was no significant correlation between the number of LVV and medial thickness in all aortic segments (Fig. 3P-T). However, the number of LVV significantly increased as the intima became thicker in Ar and I-Ab (Fig. 3K-O). There was a significant correlation between the number of LVV and intimal thickness only in the infrarenal aortic samples (Fig. 3N).

4. Discussion In this study, we identified that the distribution of VV was most dense in Ar, becoming less dense along the aorta in more distal segments, with the lowest density occurring in I-Ab. In contrast with the distribution of VV, the distribution of LVV was most dense in the infrarenal abdominal aorta. Although the two vascular structures of VV and LVV are located in proximity to each other in the adventitia, the differences of the distributions may indicate unique structural features of the aorta. With regard to VV distribution, the density of VV was correlated with aortic medial thickness, suggesting that adventitial VV were most densely distributed in Ar, becoming less dense along De, S-Ab, and I-Ab aorta. Because medial thickness is determined based on the number of elastic lamellae, the density of adventitial VV may be correlated with the number of medial elastic lamellae. Wolinsky and Glagov demonstrated that the number of elastic lamella units is correlated with blood pressure and diameter [16], thus, the correlative increase of VV density may be needed to provide sufficient nourishment to a larger and thicker aortic wall with greater radial compression. Recently, the importance of VV perfusion to maintain aortic tissue homeostasis has been reconfirmed. We have identified that adventitial VV became stenotic in the intraoperatively-harvested AAA wall and that the wall tissue was chronically ischemic and

Fig. 2. Aortic wall thickness in each segment. A: Elastica van Gieson (EVG) staining of each aortic segment. I: intima, M: media, Ar: distal arch aorta, De: descending aorta, S-Ab: suprarenal abdominal aorta, I-Ab: infrarenal abdominal aorta. Scale bars indicate 100 mm. B: Comparison of intimal thickness of each aortic segment. Data are presented as means ± standard deviations. C: Comparison of medial thickness of each aortic segment. Data are presented as means ± standard deviations. *p < 0.05, **p < 0.01.

132

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

Fig. 3. The relationship between thickness of each aortic segment and the numbers of adventitial vasa vasorum (VV) or lymphatic vasa vasorum (LVV). A-E: Relationship between intimal thickness and the number of VV. F-J: Relationship between medial thickness and the number of VV. K-O: Relationship between intimal thickness and the number of LVV. P-T: Relationship between medial thickness and the number of LVV. Ar: distal arch aorta (red), De: descending aorta (blue), S-Ab: suprarenal abdominal aorta (yellow), I-Ab: infrarenal abdominal aorta (green).

hypoxic [8]. Moreover, we showed that adventitial VV hypoperfusion and subsequent tissue hypoxia resulted in development of AAA using an animal model, in which the pathological findings of the developed aneurysms were comparable to those in human AAA

[9]. These findings strongly suggested that the sparse distribution of VV in the infrarenal abdominal aorta is associated with this segment being the most prevalent site for aortic aneurysms. In contrast, aortic dissection occurs most in Ar, and second most

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

in De, where VV are most densely distributed. More than a half century ago, Gore suggested that rupture of VV may result in intramural hematoma (IMH) following aortic dissection [8,17]. Indeed, there are many cases of aortic dissection without intimal tear [18]. Although the mechanisms of VV rupture are unknown, VV rupture-induced IMH is considered one of the major pathologies of aortic dissection [19]. Therefore, the dense distribution of VV and rich blood perfusion in arch or descending aortic wall may be associated with the prevalence of dissection in these regions. Unlike VV, adventitial LVV are most densely distributed in the infrarenal aorta. This finding seems to be particularly interesting considering that there are numerous lymphatic networks between paraaortic lymphatics and mesenteric/pelvic lymphatics in the infrarenal region. In regard to the association of lymph perfusion and aortic diseases, by using intraoperative indocyanine green fluorescence lymphography, we have previously demonstrated the presence of lymph stasis inside the infrarenal AAA wall [20,21]. Pathologically, the AAA is considered to be one of many chronic inflammatory diseases. Previous studies have shown that chronic lymph stasis is associated with tissue inflammation, it causes infiltration of mononuclear cells and tissue fibrosis [22,23]. Therefore, dense distribution of adventitial LVV and subsequent lymph stasis may be associated with aneurysmal development in the infrarenal aorta by enhancing inflammation. In this study, the number of the adventitial VV was not significantly correlated with intimal thickness throughout the aorta. However, the number of adventitial LVV of the infrarenal aorta only showed significant correlation with the intimal thickness of the aorta. Among the very limited numbers of previous reports regarding adventitial LVV [11,21,24], Drozdz et al. also showed a positive correlation between the number of adventitial LVV and intimal thickness in aortae harvested from 15 patients who were organ transplantation donors. Correlations between intimal thickness and the number of adventitial LVV have been found in aortae, carotid arteries, and iliac arteries [10,11]. Intimal thickness of the arteries associated with atherosclerotic stenosis. Interestingly, these arteries are prone to become stenotic due to atherosclerotic intimal hyperplasia. Recently, the importance of adventitial inflammation has been highlighted for its role in the pathogenesis of atherosclerosis. Adventitial inflammation with infiltration of macrophages has been reported in human atherosclerotic aorta [25]. Inflammatory cells and cytokines may be delivered via the adventitial LVV to the luminal side of vascular walls, causing intimal hyperplasia [26]. Therefore, we speculated that arteries with rich adventitial LVV may accelerate atherosclerosis. This hypothesis is compatible with the fact that atherosclerotic stenosis is most frequent in I-Ab compared to other aortic segments. However, the detailed mechanisms of atherosclerosis regarding adventitial lymph perfusionrelated inflammation remains unknown. Further studies are needed to confirm the hypothesis. The current study has limitations. First, the samples were harvested only from autopsy cases that did not have critical aortic diseases such as dissection, aneurysm, and atherosclerotic stenosis. However, the study is mainly focused on the implication to such diseases. Without comparison between samples with and without such aortic diseases, the discussion is totally speculative. Second, the small sample size is clearly limited. Third, the lack of samples from the ascending aorta is insufficient to draw conclusions regarding diseases occurring throughout the aorta. In summary, we investigated topological distributions of both adventitial VV and LVV within different segments of the aorta. The density of VV was correlated with aortic medial thickness, so that VV in I-Ab had the lowest distribution among the aortic segments;

133

however, there was no statistical correlation between adventitial VV and intimal thickness in any aortic areas. Nonetheless, we showed for the first time the distributions of adventitial LVV and their diversity from adventitial VV. Contrastingly, the density of adventitial LVV was not correlated with the aortic medial thickness. On the other hand, the distribution of LVV was most dense in I-Ab, where there was a positive correlation between the number of LVV and intimal thickness. These findings may be associated with the different prevalence of aortic diseases, such as aortic aneurysm, dissection, and atherosclerotic stenosis in each aortic area. Acknowledgements This work was supported by JSPS KAKENHI Grant number 26293310 (to NU) and JSPS KAKENHI Grant number 15K21052 (to MS). References [1] E. Allaire, F. Schneider, F. Saucy, et al., New insight in aetiopathogenesis of aortic diseases, Eur. J. Vasc. Endovasc. Surg. 37 (2009) 531e537. [2] J.C.J. Roberts, C. Moses, R.H. Wilkins, Autopsy studies in atherosclerosis I. distribution and severity of atherosclerosis in patients dying without morphologic evidence of atherosclerotic catastrophe, Circulation 20 (1959) 511e519. [3] P.G. Hagan, C.A. Nienaber, E.M. Isselbacher, et al., The international registry of acute aortic dissection (IRAD): new insights into an old disease, JAMA 283 (2000) 897e903. [4] D.D. Heistad, M.L. Marcus, Role of vasa vasorum in nourishment of the aorta, Blood Vessels 16 (1979) 225e238. [5] H. Wolinsky, S. Glagov, Nature of species differences in the medial distribution of aortic vasa vasorum in mammals, Circ. Res. 20 (1967) 409e421. [6] H.B. Benjamin, A.B. Becker, Etiologic incidence of thoracic and abdominal aneurysms, Surg. Gynecol. Obstet. 125 (1967) 1307e1310. [7] H.E. Schutte, Changes in the vasa vasorum of the atherosclerotic aortic wall, Angiologica 5 (1968) 210e222. [8] H. Tanaka, N. Zaima, T. Sasaki, et al., Adventitial vasa vasorum arteriosclerosis in abdominal aortic aneurysm, PLoS One 8 (2013) e57398. [9] H. Tanaka, N. Zaima, T. Sasaki, et al., Hypoperfusion of the adventitial Vasa Vasorum develops an abdominal aortic aneurysm, PLoS One 10 (2015) e0134386. [10] K. Drozdz, D. Janczak, P. Dziegiel, et al., Adventitial lymphatics of internal carotid artery in healthy and atherosclerotic vessels, Folia Histochem Cytobiol. 46 (2008) 433e436. [11] K. Drozdz, D. Janczak, P. Dziegiel, et al., Adventitial lymphatics and atherosclerosis, Lymphology 45 (2012) 26e33. [12] S. Breiteneder-Geleff, A. Soleiman, R. Horvat, et al., Podoplaninea specific marker for lymphatic endothelium expressed in angiosarcoma, Verh. Dtsch. Ges. Pathol. 83 (1999) 270e275. [13] N.G. Ordonez, Immunohistochemical endothelial markers: a review, Adv. Anat. Pathol. 19 (2012) 281e295. [14] S. Banerji, J. Ni, S.X. Wang, et al., LYVE-1, a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for hyaluronan, J. Cell Biol. 144 (1999) 789e801. [15] N.G. Ordonez, Value of podoplanin as an immunohistochemical marker in tumor diagnosis: a review and update, Appl. Immunohistochem. Mol. Morphol. 22 (2014) 331e347. [16] H. Wolinsky, S. Glagov, Comparison of abdominal and thoracic aortic medial structure in mammals. Deviation of man from the usual pattern, Circ. Res. 25 (1969) 677e686. [17] I. Gore, Pathogenesis of dissecting aneurysm of the aorta, AMA Arch. Pathol. 53 (1952) 142e153. [18] K.J. Macura, F.M. Corl, E.K. Fishman, et al., Pathogenesis in acute aortic syndromes: aortic dissection, intramural hematoma, and penetrating atherosclerotic aortic ulcer, AJR Am. J. Roentgenol. 181 (2003) 309e316. [19] R. Erbel, F. Alfonso, C. Boileau, et al., Diagnosis and management of aortic dissection, Eur. Heart J. 22 (2001) 1642e1681. [20] N. Unno, K. Inuzuka, M. Suzuki, et al., Preliminary experience with a novel fluorescence lymphography using indocyanine green in patients with secondary lymphedema, J. Vasc. Surg. 45 (2007) 1016e1021. [21] M. Sano, T. Sasaki, S. Hirakawa, et al., Lymphangiogenesis and angiogenesis in abdominal aortic aneurysm, PLoS One 9 (2014) e89830. [22] M. Sugaya, Y. Kuwano, H. Suga, et al., Lymphatic dysfunction impairs antigenspecific immunization, but augments tissue swelling following contact with allergens, J. Invest Dermatol. 132 (2012) 667e676. [23] J.C. Zampell, S. Aschen, E.S. Weitman, et al., Regulation of adipogenesis by lymphatic fluid stasis: part I. Adipogenesis, fibrosis, and inflammation, Plast. Reconstr. Surg. 129 (2012) 825e834.

134

M. Sano et al. / Atherosclerosis 247 (2016) 127e134

[24] C.A. Dutertre, M. Clement, M. Morvan, et al., Deciphering the stromal and hematopoietic cell network of the adventitia from non-aneurysmal and aneurysmal human aorta, PLoS One 9 (2014) e89983. [25] P.R. Moreno, K.R. Purushothaman, V. Fuster, et al., Intimomedial interface damage and adventitial inflammation is increased beneath disrupted

atherosclerosis in the aorta: implications for plaque vulnerability, Circulation 105 (2002) 2504e2511. [26] X. Xu, H. Lu, H. Lin, et al., Lymphangiogenesis promotes inflammation and neointimal hyperplasia after adventitia removal in the rat carotid artery, Int. J. Cardiol. 134 (2009) 426e427.