Extracellular matrix dynamics associated with tissue-engineered intravascular sclerotherapy

Journal of Pediatric Surgery (2006) 41, 757 – 762

www.elsevier.com/locate/jpedsurg

Independent original articles

Extracellular matrix dynamics associated with tissue-engineered intravascular sclerotherapy Adam M. Vogela,e, C. Jason Smithersa,e, Harry P. Kozakewichb,e, David Zurakowskic, Marsha A. Mosesa, Patricia E. Burrowsd,e, Dario O. Fauzaa, Steven J. Fishmana,e,* a

Department of Surgery, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115, USA Department of Pathology, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115, USA c Department of Orthopaedics, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115, USA d Department of Radiology, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115, USA e Vascular Anomalies Center, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115, USA b

Index words: Tissue-engineering; Venous malformation; Sclerotherapy; Extracellular matrix; Matrix metalloproteinases; Tissue inhibitors of matrix metalloproteinases

Abstract Background: The extracellular dynamics after intravascular sclerotherapy with an injectable, fibroblastbased engineered construct is unknown. Methods: Rabbits underwent ethanol sclerotherapy of a jugular vein segment. Control animals (n = 40) underwent no further treatment or an acellular collagen hydrogel was injected. Experimental animals (n = 20) received a tissue-engineered construct. After 1, 2, 4, and 20 to 24 weeks, segments were evaluated for collagen, glycosaminoglycan (GAG), matrix metalloproteinase (MMP) 2 and 9, and tissue inhibitors of MMP (TIMPs) 1 and 2 and scored on a scale of 0 to 3. Groups and time points were compared using nonparametric statistical analysis. Results: Collagen content was higher in animals that received fibroblasts ( P b .05). Glycosaminoglycan analysis showed a higher grade only at 1 week ( P b .05). Collagen and GAG deposition were prominent at weeks 1 through 4, and decreased over time. Both MMP-2 and MMP-9 and TIMP-1 and TIMP-2 grade decreased with time ( P b .01) in all groups, with no differences between groups. Conclusion: Enhancement of intravascular sclerotherapy by tissue engineering stems, at least in part, from increased local deposition of collagen and GAG. MMP and TIMPs may play a role in recanalization after experimental sclerotherapy. Tissue engineering may be a valuable adjunct for the treatment of vascular malformations. D 2006 Elsevier Inc. All rights reserved.

Venous malformations are congenital lesions of dysmorphogenesis that cause significant morbidity and mortality * Corresponding author. Department of Surgery, Children’s Hospital Boston, Boston, MA 02115, USA. Tel.: +1 617 355 3040; fax: +1 617 730 0477. E-mail address: [email protected] (S.J. Fishman). 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2006.02.021

[1]. Luminal recanalization through thrombus organization after sclerotherapy is responsible for treatment failure after intravascular sclerotherapy [2,3]. The process of recanalization is mediated through the various cellular and extracellular matrix (ECM) elements and is analogous to wound healing [4]. Collagen and glycosaminoglycans (GAGs) are the main structural elements of ECM and are

758 influenced by matrix metalloproteinases (MMPs) and tissue inhibitors of MMPs (TIMPs). In addition, MMP and TIMPs are essential for vascular remodeling after injury and thrombosis [5-8]. These molecules also predict the extent and progression of vascular anomalies [9]. We have previously shown in a leporine model of venous malformations that the injection of a tissue-engineered construct consisting of autologous fibroblasts in a collagen-based hydrogel carrier after ethanol sclerotherapy results in a significant increase in long-term angiographic vascular occlusion. In this analysis, we explore the relationships between the injected tissue-engineered construct and ECM dynamics.

1. Materials and methods The Harvard Medical School animal management program is sanctioned by the American Association for the Accreditation of Laboratory Animal Care and meets National Institutes of Health standards as set forth in the Guide for the Care and Use of Laboratory Animals. The Harvard Medical School Standing Committee on Animals approved the present study, under protocol no. 03583. This study represents the analysis of pathologic specimens acquired during previously reported experiments [2,3]. Briefly, intraluminal sclerotherapy with dehydrated alcohol was performed on an isolated external jugular vein segment through cervicotomy under general anesthesia in

A.M. Vogel et al. New Zealand white rabbits (n = 40). For the purposes of the following pathologic analysis, animals were divided into 2 groups. In the control or acellular group (n = 20), either no further manipulations were performed or collagen hydrogel was injected. In the experimental or cellular group (n = 15), collagen hydrogel seeded with autologous fibroblasts was injected. Animals underwent euthanasia at 1, 2, 4, and 20 to 24 weeks postoperatively. Manipulated vein segments were stained for collagen using trichrome and for GAGs with alcian blue using standard methods. Immunohistochemistry was performed using 5-lm-thick formalin-fixed, paraffinembedded tissue sections with standard laboratory techniques and using antibodies against for MMP-2 and MMP-9 (Chemicon International Inc, Temecula, Ca) and TIMP-1 and TIMP-2 (EMD Biosciences, San Diego, Calif) [8,10]. Individual slides were graded from 0 to 3 according to intensity and overall distribution. Slides were labeled with an animal identification number that did not correspond to the experimental or control group. Outcome variables were graded at 1, 2, 4, and 24 weeks and expressed in terms of the median and interquartile range owing to the small sample sizes at each time point and the nonnormality of the data [11]. Histopathology grading for collagen, GAG, MMP-2, MMP-9, TIMP-1, and TIMP-2 were compared between the experimental and control groups using the nonparametric Mann-Whitney rank-sum test [12]. Differences in grading between week 1 and each of the other time points were assessed using the Mann-Whitney

Fig. 1 Representative histology illustrating trichrome collagen staining grades 0 through 3. Grade 0 (A) shows an overall paucity of collagen staining in the vascular lumen. Minimal collagen fibers are present within the vascular wall. Grade 1 specimens (B) show deposition along the periphery of the intraluminal thrombus, generally in association with invading capillaries or inflammatory cells. Grade 2 specimens (C) show modest collagen deposition associated with invading neovasculature and involve a significant portion of the vascular lumen. Collagen deposition appears in a centripetal manner from the periphery of the vessel inward, paralleling the growth of capillaries from the vessel wall and overall thrombus organization. Grade 3 specimens (D) demonstrate collagen deposition throughout the intraluminal thrombus. (E-H) Representative histology of alcian blue GAG staining grades 0 through 3. Grade 0 (E) demonstrates an absence of GAG staining in the vascular lumen. Grade 1 specimens (F) show GAG deposition (light blue stain) along the periphery of the intraluminal thrombus. Grade 2 specimens (G) show more intense GAG deposition involving a significant portion of the vascular lumen. Grade 3 specimens (H) demonstrate GAG deposition throughout the organizing thrombus. In general, GAG deposition also appears in a centripetal manner from the periphery of the vessel inward in parallel with capillary growth.

ECM dynamics and intravascular sclerotherapy rank-sum test with summary data presented in terms of the median grade. Statistical analysis was performed using the SPSS statistical package (version 12.0, SPSS Inc, Chicago, Ill). All P values are 2-tailed using a significance level of .05 for rejecting the null hypothesis.

2. Results 2.1. Collagen Representative sections of trichrome staining grades 0 through 3 are shown in Fig. 1A-D. Collagen content was most prominent at weeks 1 through 4 and, in general, diminished over time as illustrated in Fig. 2. In addition, at each time point, collagen content was significantly higher in animals that received fibroblasts as seen in Table 1 ( P b .05; Mann-Whitney U test).

2.2. Glycosaminoglycans Representative sections of alcian blue staining grades 0 through 3 are shown in Fig. 1E-H. Like collagen, GAG content was more prominent at weeks 1 through 4 and diminished over time (Fig. 2). Comparison of GAG content between experimental and control groups showed a

759 Table 1

Extracellular matrix components Experimental group

Control group

P

Collagen (wk) 1 2 (1.5-2) 2 2 (2-2.5 4 3 (1.5-3) 24 1 (1-1)

1 1 1 0

(1-1) (0.75-2) (0-1.25) (0-1)

b.05a b.05a b.05a b.05a

GAG (wk) 1 2 (1.5-2.5) 2 2 (1-2.5) 4 1 (1-1.75) 24 1 (0-1)

1 1 1 0

(1-1) (1-2) (0.75-2) (0-1)

b.05a .31 .84 .43

Data represent the median grade with the interquartile range shown in parentheses. a Statistically significant.

significantly higher grade only at 1 week ( P b .05; MannWhitney U test) (Table 1).

2.3. Matrix metalloproteinase 2 and 9 Representative sections of immunohistochemical staining of MMP-2 (grade 0-3) can be found in Fig. 3A-D. Representative histology of MMP-9 immunostaining grades 0 through 3 can be found in Fig. 3E-H. There were no significant differences in MMP-2 and MMP-9 between groups (Table 2). However, significant time-related decreases in grade were seen in MMP-2 and MMP-9 staining as illustrated in Fig. 4 ( P b .01; Kruskal-Wallis test).

2.4. Tissue inhibitor of MMP-1 and MMP-2 Representative sections of immunohistochemical staining of TIMP-1 (grade 0 through 2) are demonstrated in Fig. 5A-C. Staining of TIMP-2 (grade 0 through 2) can be found in Fig. 5D-F. There were no significant differences in TIMP-1 and TIMP-2 between groups (Table 2). However, as with MMP-2 and MMP-9, significant time-related decreases in grade were seen in TIMP-1 and TIMP-2 staining as illustrated in Fig. 6 ( P b .01; Kruskal-Wallis test).

3. Discussion

Fig. 2 The progression of collagen and GAG grades over time. *P b .05 indicates significant difference in grade when compared with the initial (week 1) time point.

The recurrence of symptoms after intravascular sclerotherapy for the treatment of venous malformations is thought to be mediated by intraluminal thrombus organization and recanalization [3]. The physiologic response of blood vessels after intraluminal sclerotherapy and acute thrombus formation parallels wound healing [4]. A welldescribed temporal and topographical sequence of cellular trafficking directs thrombus organization [13-16]. Neutrophils, followed by macrophages, migrate into the thrombus through the vein wall in concert with vascular invasion [16-18]. Macrophages promote thrombus organization by releasing several angiogenic factors [6,13,19,20]. Finally,

760

A.M. Vogel et al.

Fig. 3 Representative histology of MMP-2 immunostaining grades 0 through 3 (A-D) and MMP-9 immunostaining grades 0 through 3 (E-H). Dark brown color represents positive staining. In general, both MMP-2 and MMP-9 localize to areas with more prominent concentrations of inflammatory cells such as macrophages and multinucleated giant cells.

infiltrating fibroblasts depositing ECM proceeds in a similar pattern [21]. These biochemical and cellular events are modulated by MMPs and their inhibitors [4,6,22]. Matrix metalloproteinases represent a series of structurally related enzymes whose common function is to degrade ECM components and are released after vascular injury [23-25]. MMP-2 and MMP-9 (gelatinase A and B, respectively) are important in collagen maturation, interstitial collagen matrix degradation, Table 2

endothelial basement membrane degradation, and angiogenesis [23]. Mononuclear cells use metalloproteinases to overcome thrombus and ECM barriers [24,26]. Extracellular proteolysis by MMPs is held in check by inhibitor molecules [25]. The 4 known TIMPs expressed by

MMP and TIMP Experimental group

Control group

P

MMP-2 (wk) 1 2 (1.5-2.5) 2 3 (1.5-3) 4 2 (2-2.75) 24 0.5 (0-1)

2 2 1 0

(2-2.25) (2-2.25) (1-3) (0-1)

.86 .51 .54 .80

MMP-9 (wk) 1 2 (1.5-3) 2 2 (1.5-3) 4 1.5 (1-2.75) 24 0.5 (0-1)

1 2 1 0

(1-2) (2-3) (1-2) (0-1)

.17 .95 .37 .56

TIMP-1 (wk) 1 1 (0-1) 2 1 (0.5-1.5) 4 1.5 (1-2) 24 0 (0-0.25)

0 1 1 0

(0-1) (1-1) (1-2) (0-1)

.59 .99 .64 .49

TIMP-2 (wk) 1 1 (0.5-2) 2 1 (1-2) 4 2 (1.25-2) 24 0 (0-1)

1 1 1 0.5

(1-2) (1-2) (1-2) (0-1)

.95 .98 .37 .64

Data represent the median grade with the interquartile range shown in parentheses.

Fig. 4 The progression of MMP-2 and MMP-9 grades over time. *P b .05 indicates significant difference in grade when compared with initial (week 1) time point.

ECM dynamics and intravascular sclerotherapy

761

Fig. 5 Representative histology illustrating TIMP-1 immunostaining grades 0 through 2 (A-C) and TIMP-2 immunostaining grades 0 through 3 (D-F). TIMP-1 and TIMP-2 (dark brown) localize to areas with prominent concentrations of inflammatory cells.

vascular endothelium combine in a one-to-one ratio with MMP [24]. Tissue inhibitor of MMP-1 inhibits MMP-9 and TIMP-2 inhibits MMP-2 [27]. Organization of arterial thrombi is associated with stable expression of MMP-2 and stable or decreased expression of TIMP-1 and TIMP-2, which may help shift extracellular proteolysis toward thrombus resolution [26]. The spatial pattern and density of MMP immunostaining observed in this experiment are consistent with previous experiments demonstrating MMP presence at sites of inflammation and angiogenesis. There were no differences between in MMP or TIMP grade between the experimental and control groups. This may be explained by the fact that both groups were exposed to identical intraluminal injuries, which resulted in similar inflammatory responses [2,3]. Although the presence of additional fibroblasts increased the density of ECM structural components, they did not appear to impact the magnitude of the inflammatory response. The increased ECM seen is likely a result of greater cellular substrate to produce matrix responding to identical stimulation. Thus, an appropriately proportional increase in matrix production may tip the physiologic balance toward more permanent vascular occlusion after traditional sclerotherapy [3]. Other angiogenic molecules such as vascular endothelial growth factor and basic fibroblast growth factor impact MMP function [25,28,29]. Thrombin, plasmin, and plasminogen also may promote overall MMP activation at sites of vascular injury [6,20,27]. Additional inflammatory mediators such as tumor necrosis factor, interleukin 8, and monocyte chemotactic protein 1 are important mediators of inflammation, neovascularization, and subsequent thrombus maturation [13,14,18,21,30]. Similarly, the up-regulation of early adhesion molecules after vein injury helps coordinate inflammatory cell infiltration and the general inflammatory and organization process [15,31]. Additional experiments will be necessary to characterize the role of these variables in this model.

There are several limitations to the present study. We compared ECM composition in sclerosed veins with or without the addition of a tissue-engineered, fibroblast-based construct. Although it would not be likely to change our conclusions regarding the experimental and control groups, no analysis was performed with respect to the normal vessel wall components. Similarly, histologic and immunohistochemical

Fig. 6 TIMP-1 and TIMP-2 grade progression over time. *P b .05 indicates significant difference in grade when compared with the initial (week 1) time point.

762 analysis of ECM structural and regulatory molecules, respectively, was based on subjective interpretation. Although this type of qualitative analysis is well described in the literature, function data, such as quantitative matrix analysis and tissue zymography, would be helpful in better describing the ECM dynamics present in this model. In conclusion, our analysis demonstrates that an injected tissue-engineered, fibroblast-based autologous construct after ethanol sclerotherapy results in increased ECM deposition over time. This process is likely mediated through a proportionally larger fibroblast response to the local inflammatory milieu produced by endothelial chemical injury and regulated by MMPs and their inhibitors. Additional experiments are necessary to further quantify this response and the underlying molecular mediators to more effectively tailor this system for future clinical use in the treatment of venous malformations.

References [1] Mulliken JB, Fishman SJ, Burrows PE. Vascular anomalies. Curr Probl Surg 2000;37:517 - 84. [2] Smithers CJ, Vogel AM, Kozakewich HPW, et al. Enhancement of intravascular sclerotherapy by tissue engineering: short term results. J Pediatr Surg [in press]. [3] Smithers CJ, Vogel AM, Kozakewich HP, et al. An injectable tissueengineered embolus prevents luminal recanalization after vascular sclerotherapy. J Pediatr Surg [in press]. [4] Singer AJ, Clark RA. Cutaneous wound healing. N Engl J Med 1999;341:738 - 46. [5] Gong YL, Xu GM, Huang WD, et al. Expression of matrix metalloproteinases and the tissue inhibitors of metalloproteinases and their local invasiveness and metastasis in Chinese human pancreatic cancer. J Surg Oncol 2000;73:95 - 9. [6] Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res 2002;90: 251 - 62. [7] Kuzuya M, Iguchi A. Role of matrix metalloproteinases in vascular remodeling. J Atheroscler Thromb 2003;10:275 - 82. [8] Feldman LJ, Mazighi M, Scheuble A, et al. Differential expression of matrix metalloproteinases after stent implantation and balloon angioplasty in the hypercholesterolemic rabbit. Circulation 2001; 103:3117 - 22. [9] Marler JJ, Fishman SJ, Kilroy SM, et al. Increased expression of urinary matrix metalloproteinases parallels the extent and progression of vascular anomalies. Pediatrics [in press]. [10] Feldman LJ, Aguirre L, Ziol M, et al. Interleukin-10 inhibits intimal hyperplasia after angioplasty or stent implantation in hypercholesterolemic rabbits. Circulation 2000;101:908 - 16. [11] Armitage P, Berry G, Matthews JNS. Statistical methods in medical research. 4th ed. Vol. 2002, Oxford Williston (VT): Blackwell Science; Distributor, Blackwell Publishing c/o AIDC; 2002. xi, p. 817. [12] Glantz SA. Primer of biostatistics. 5th ed. Vol. 2001, New York: McGraw-Hill, Medical Pub. Div; 2001. xviii, p. 489.

A.M. Vogel et al. [13] Wakefield TW, Linn MJ, Henke PK, et al. Neovascularization during venous thrombosis organization: a preliminary study. J Vasc Surg 1999;30:885 - 92. [14] Wakefield TW, Greenfield LJ, Rolfe MW, et al. Inflammatory and procoagulant mediator interactions in an experimental baboon model of venous thrombosis. Thromb Haemost 1993;69:164 - 72. [15] Downing LJ, Strieter RM, Kadell AM, et al. IL-10 regulates thrombus-induced vein wall inflammation and thrombosis. J Immunol 1998;161:1471 - 6. [16] Scott GB. A quantitative study of the fate of occlusive red venous thrombi. Br J Exp Pathol 1968;49:544 - 50. [17] Sevitt S. The mechanisms of canalisation in deep vein thrombosis. J Pathol 1973;110:153 - 65. [18] Henke PK, Wakefield TW, Kadell AM, et al. Interleukin-8 administration enhances venous thrombosis resolution in a rat model. J Surg Res 2001;99:84 - 91. [19] Moldovan NI, Asahara T. Role of blood mononuclear cells in recanalization and vascularization of thrombi: past, present, and future. Trends Cardiovasc Med 2003;13:265 - 9. [20] Henke PK, Varga A, De S, et al. Deep vein thrombosis resolution is modulated by monocyte CXCR2-mediated activity in a mouse model. Arterioscler Thromb Vasc Biol 2004;24:1130 - 7. [21] Humphries J, McGuinness CL, Smith A, et al. Monocyte chemotactic protein–1 (MCP-1) accelerates the organization and resolution of venous thrombi. J Vasc Surg 1999;30:894 - 9. [22] Geary RL, Nikkari ST, Wagner WD, et al. Wound healing: a paradigm for lumen narrowing after arterial reconstruction. J Vasc Surg 1998;27:96 - 106 [discussion 106-8]. [23] Whatling C, McPheat W, Hurt-Camejo E. Matrix management: assigning different roles for MMP-2 and MMP-9 in vascular remodeling. Arterioscler Thromb Vasc Biol 2004;24:10 - 1. [24] Chase AJ, Newby AC. Regulation of matrix metalloproteinase (matrixin) genes in blood vessels: a multi-step recruitment model for pathological remodelling. J Vasc Res 2003;40:329 - 43. [25] Pepper MS. Role of the matrix metalloproteinase and plasminogen activator–plasmin systems in angiogenesis. Arterioscler Thromb Vasc Biol 2001;21:1104 - 17. [26] Raymond J, Lebel V, Ogoudikpe C, et al. Recanalization of arterial thrombus, and inhibition with beta-radiation in a new murine carotid occlusion model: MRNA expression of angiopoietins, metalloproteinases, and their inhibitors. J Vasc Surg 2004;40:1190 - 8. [27] Galis ZS, Kranzhofer R, Fenton II JW, et al. Thrombin promotes activation of matrix metalloproteinase-2 produced by cultured vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 1997; 17:483 - 9. [28] Fang J, Shing Y, Wiederschain D, et al. Matrix metalloproteinase–2 is required for the switch to the angiogenic phenotype in a tumor model. Proc Natl Acad Sci U S A 2000;97:3884 - 9. [29] Burbridge MF, Coge F, Galizzi JP, et al. The role of the matrix metalloproteinases during in vitro vessel formation. Angiogenesis 2002;5:215 - 26. [30] Wakefield TW, Strieter RM, Wilke CA, et al. Venous thrombosisassociated inflammation and attenuation with neutralizing antibodies to cytokines and adhesion molecules. Arterioscler Thromb Vasc Biol 1995;15:258 - 68. [31] Downing LJ, Wakefield TW, Strieter RM, et al. Anti–P-selectin antibody decreases inflammation and thrombus formation in venous thrombosis. J Vasc Surg 1997;25:816 - 27 [discussion 828].