Histochem Cell Biol DOI 10.1007/s00418-015-1317-3
ORIGINAL PAPER
Neointimal hyperplasia in allogeneic and autologous venous grafts is not different in nature Busch Albert1 · Hartmann Elena2 · Wagner Nicole3 · Ergün Süleyman3 · Kickuth Ralph4 · Kellersmann Richard1 · Lorenz Udo1
Accepted: 3 March 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Neointimal hyperplasia, transplant rejection and thus immunogenicity of allografts are possible reasons for poorer patency rates in cryopreserved venous allografts for peripheral bypass surgery in comparison with autologous venous grafts. To expand the limited knowledge from human allografts, we histologically investigated allogeneic and autologous venous grafts in arterial location. Specimens of allogeneic and autologous venous graft stenosis, harvested 6 months after bypass implantation, were immunohistochemically characterized. Examination of the lesions showed a uniform morphological pattern. A continuous endothelial layer, tissue fibrosis and a thickened neointima with monocytes and dedifferentiated vascular smooth muscle cells were seen in both conduits with very low cell turnover and the absence of acute and chronic inflammation. Neoangiogenesis with CD34-positive endothelium was abundant in the vessel media. The morphological patterns of allogeneic and autologous neointima formation are similar. Consequently, neointimal hyperplasia in venous grafts Electronic supplementary material The online version of this article (doi:10.1007/s00418-015-1317-3) contains supplementary material, which is available to authorized users. * Busch Albert [email protected] 1
Department for General, Visceral, Vascular and Paediatric Surgery, University Clinic of Würzburg – ZOM, Oberduerrbacher Str 6, 97080 Würzburg, Germany
2
Institute of Pathology, University of Würzburg, Würzburg, Germany
3
Institute of Anatomy and Cell Biology, University of Würzburg, Würzburg, Germany
4
Institute of Diagnostic and Interventional Radiology, University of Würzburg, Würzburg, Germany
may reflect a uniform physiological host response of nonimmunological factors with the reasons for poorer clinical outcome of cryopreserved allografts yet to be elucidated. Keywords Cryopreserved vein · Bypass stenosis · Neointimal hyperplasia · Angiogenesis · Chronic limb ischemia Abbreviations PAOD Peripheral arterial disease PTFE Polytetrafluoroethylene VSMC Vascular smooth muscle cell eNOS Endothelial nitric oxide synthase ECM Extracellular matrix VV Vasa vasorum VEGF Vascular endothelial growth factor DSA Digital subtraction angiography DAPI 4,6-Diamidin-2-phenylindol EC Endothelial cell CD Cluster of differentiation PECAM1 Platelet/endothelial cell adhesion molecule
Introduction Autologous vein is the favored conduit for bypass revascularization in chronic critical limb ischemia (CLI) regarding long-time patency and infection control (Conte 2010). If the saphenous or arm veins are absent or not eligible for surgery, prosthetic grafts made from polytetrafluoroetylene (PTFE) or polyester (Dacron®) are commonly used alternatives bearing inferior patency rates and the risk of graft infection. Allografts such as cryopreserved cadaveric veins are therefor considered a further alternative since 1976 (Harris et al. 2001).
13
Although the biological effects of cryoconservation on complex tissues such as a human vessel concerning reactivity, immunology and thrombogenicity are still elusive, the process itself has been well established and allografts are commercially available (Fahner et al. 2006; Muller-Schweinitzer et al. 2007). Feasibility in cardiac and peripheral bypass procedures as well as vascular access has been demonstrated, and clinical series report safe utilization in infectious and non-infectious sites (Faggioli and Ricotta 1994). One-year primary patency rates between 17 and 40 % for infrapopliteal reconstructions, however, emphasize clear indications regarding outcome, costs and frequency of subsequent interventions (Barshes et al. 2013; Farber et al. 2003; Harris et al. 2001; Martin et al. 1994; Walker et al. 1993). Arterialization of venous grafts is reflected by neointimal hyperplasia (NH), but may occur excessively or uncontrolled at the proximal anastomosis site, resulting in highly rigid lesions requiring aggressive treatment such as cutting balloon angioplasty or surgery enabling secondary patency rates of 45–63 % at 12 months (Kreienberg et al. 2007; Leseche et al. 1997; Zehr et al. 2011). Various stimuli concert NH. Endothelial denudation with exposure of procoagulant factors, platelet adhesion and leukocyte extravasation, vasospasm due to denervation and hypoxia, triggered by loss of vasa vasorum, mediates an acute injury response with consecutive extracellular matrix (ECM) degradation and monocyte homing (Collins et al. 2012; Davies and Hagen 2011). Antigenicity of allografts may further result in rapid initiation and progression of NH. Phenotype switch in vascular smooth muscle cell (VSMC), the main cells of the tunica media, from a quiescent contractile to a dedifferentiated proliferative type with migratory potential appears to be the main downstream effector (Lacolley et al. 2012). Additional loss of eNOSdriven endothelial relaxation response leads to a thickened, rigid neointima, predestined at sites with high turbulent flow and increased shear stress, such as the proximal anastomosis with proximity of host and graft tissue (Fig. 1; Collins et al. 2012; Muto et al. 2010). Angiogenesis within the tunica media is frequently seen in atherosclerotic plaque formation or venous grafts in coronary or peripheral position (Khurana et al. 2005; Westerband et al. 2000). Exact trigger mechanisms within the complex network of pro- and anti-angiogenic factors remain elusive. Expression of, i.e., VEGF due to tissue hypoxia might be of importance (Collins et al. 2012). Angiogenesis may as well be part of chronic vascular transplant rejection, but is, however, evident in NH. Various pharmaceutical and mechanical strategies have been assessed to improve outcome after autologous or allogeneic bypass procedures. Handling of the surgically harvested vein is a strong outcome determinant (Conte 2009;
13
Histochem Cell Biol
Santo et al. 2013). Preimplantation histological analysis has been proposed (Giannoukas et al. 1997). External sheathing and brachytherapy failed to prevent NH (Kohler et al. 1989). Promising results have come from topical drug application culminating in the PREVENT III trial, administering edifoligide, a VSMC modulator, which in the end did not improve outcome and therefor highlights the incomplete understanding of the pathways involved (Wiedemann et al. 2012). The cytostatic agents rapamycin and paclitaxel showed short-term inhibition but did not prevent NH after a total of 12 weeks (Murphy et al. 2007; Shukla and Jeremy 2012). The anti-angiogenic drug thalidomide attenuates experimental atherosclerosis but has never been used in intimal proliferation models. In summary, many strategies have been used to prevent NH, a phenomenon incompletely understood and supposedly in parts physiological, in a heterogeneous clinical setting. Thus, it seemed pertinent to further evaluate the morphological characteristics of NH in different human grafts. We present for the first time a detailed histological analysis of cryopreserved venous allograft stenosis due to NH in comparison with autologous bypass grafts 6 months after implantation.
Materials and methods Patients’ characteristics Stenosis samples were harvested during open surgical bypass repair 6 months after femorocrural revascularization with an ABO but not sex-matched cryopreserved human greater saphenous vein (CryoLife©) (specimen 1 and 2) or after an autologous greater saphenous vein femorocrural bypass (specimen 3 and 4), respectively. Control samples are of greater saphenous vein. Acquisition was in accordance with the Declaration of Helsinki, the local ethic review committee and with patients’ informed and written consent. Indications for revascularization procedures were CLI (patient 2) or tissue loss (patients 1, 3 and 4) with lack of autologous vein. Open bypass repair was due to recurrent rest pain and impaired lesion healing due to stenosis at the proximal anastomotic site seen on angiography (Fig. 1). All four patients had good arterial inflow at the time of operation and at least one patent crural target vessel with good pedal runoff. Bypasses were in reversed position without valvulotomy maneuver (see supplementary Table 1). Specimen acquisition and histochemistry Tissue samples were excised during operation and immediately fixated in 3, 5 % formaldehyde (Fischer,
Histochem Cell Biol
Fig. 1 Digital subtraction angiography image (a) and intraoperative situs photography (b) reveal a circumferential rigid stenosis (white arrow) with intimal thickening of the femorocrural cryopreserved vein allograft bypass of patient 2 at the proximal anastomosis at the femoral bifurcation (diamond), 6 months after implantation. The profunda femoris artery (asterisk) shows strong collateralization but only few calcifications. HE-stained sections of stenotic lesions of cryopreserved allografts (c, d patients 1, 2) and autologous vein grafts (f, g, patients 3, 4): all four stenotic specimen show a similar morphology distinct from the control specimen (e). A neointima (N) rich in extra-
cellular matrix is absent in the control vein. Here, the tunica media (M) consists of a collagen and elastin meshwork and vascular smooth muscle cells with a longitudinal portion luminal and a circular portion orientated toward the tunica adventitia (A) which has lost its orientation in the stenotic lesions (c, d, f, g). Small vessels (25 ± 11 µm diameter) spreading into the tunica media can be depicted in every graft sample but are completely absent in control samples. Pinkstained collagen is absent in the neointima of stenotic lesions as are elastic fibers. (Scale bar 100 µm, magnification ×40)
Table 1 List of all primary antibodies used in the study by target, identification number, manufacturer and dilution used in our experiments
Germany). One-micrometer samples were mounted on Superfrost© slides (Menzel, Germany) and primarily stained for hematoxylin and eosin (HE) and elastica van Gieson (EvG).
Target
Antibody
Manufacturer
Dilution
Ki67 CD31 CD34 Actin CD3 CD20 vWF CD68 Vimentin Desmin
ab16667 M0823 M7165 ab5694 ab16669 ab9475 F-3520 M0814 ab8978 ab32362
Abcam Dako Dako Abcam Abcam Abcam Sigma-Aldrich Dako Abcam Abcam
1:50 1:100 1:100 1:200 1:200 1:40 1:1000 1:100 1:200 1:100
Kiel Institute of Pathology
1:40,000
KiM1P
KiM1P is not commercially available, but has been used and described for many years in daily routine as well as experimental studies by the Institute for Pathology, University of Wuerzburg (further information is available upon request to the author)
Immunohistochemistry Table 1 shows the different antibodies used. Counterstaining was done with either nuclear fast red aluminum or Mayer’s hematoxylin (both from Carl Roth, Germany). Fluorescence Immunohistochemistry For double immunofluorescence, CD31 (ab28364, rabbit polyclonal, Abcam) and CD34 (M7165, mouse monoclonal, Dako) were conjugated with Cy3- and Cy5-labeled secondary antibodies (both from KPL, US). Counterstaining was DAPI (LifeTechnologies®).
13
Light microscopy Light microscopy images were made by Keyence BZ9000 microscope (Keyence, Japan). Image editing as well as size measurements within the samples was performed with AZ Analyser II software distributed by Keyence, Japan. Confocal laser scanning microscopy Confocal laser scanning microscopy images were taken using a Nikon Eclipse TI-E inverse confocal microscope equipped with Coherent sapphire lasers and an mbp communication visible fiber laser (lines: 405, 488, 561, 647 nm). Objective (Plan Apo 60x NA:1, 4) was used and detection in three simultaneous channels applied. The system was equipped with NIS-Elements Advanced Research Software (Nikon). ImageJ (1.47v, NIH, USA http://magej. nih.gov/ij) was used for editing of confocal laser scanning microscope pictures.
Results Microscopic and immunohistochemical examination of stenotic lesions revealed a uniform morphological pattern of autologous bypass and cryopreserved allograft tissue with distinct features compared with control vein specimen (Fig. 1). The neointima, lying between the endothelium and the lamina elastica interna, was between 142 and 760 µm thick (normal intima
ORIGINAL PAPER
Neointimal hyperplasia in allogeneic and autologous venous grafts is not different in nature Busch Albert1 · Hartmann Elena2 · Wagner Nicole3 · Ergün Süleyman3 · Kickuth Ralph4 · Kellersmann Richard1 · Lorenz Udo1
Accepted: 3 March 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Neointimal hyperplasia, transplant rejection and thus immunogenicity of allografts are possible reasons for poorer patency rates in cryopreserved venous allografts for peripheral bypass surgery in comparison with autologous venous grafts. To expand the limited knowledge from human allografts, we histologically investigated allogeneic and autologous venous grafts in arterial location. Specimens of allogeneic and autologous venous graft stenosis, harvested 6 months after bypass implantation, were immunohistochemically characterized. Examination of the lesions showed a uniform morphological pattern. A continuous endothelial layer, tissue fibrosis and a thickened neointima with monocytes and dedifferentiated vascular smooth muscle cells were seen in both conduits with very low cell turnover and the absence of acute and chronic inflammation. Neoangiogenesis with CD34-positive endothelium was abundant in the vessel media. The morphological patterns of allogeneic and autologous neointima formation are similar. Consequently, neointimal hyperplasia in venous grafts Electronic supplementary material The online version of this article (doi:10.1007/s00418-015-1317-3) contains supplementary material, which is available to authorized users. * Busch Albert [email protected] 1
Department for General, Visceral, Vascular and Paediatric Surgery, University Clinic of Würzburg – ZOM, Oberduerrbacher Str 6, 97080 Würzburg, Germany
2
Institute of Pathology, University of Würzburg, Würzburg, Germany
3
Institute of Anatomy and Cell Biology, University of Würzburg, Würzburg, Germany
4
Institute of Diagnostic and Interventional Radiology, University of Würzburg, Würzburg, Germany
may reflect a uniform physiological host response of nonimmunological factors with the reasons for poorer clinical outcome of cryopreserved allografts yet to be elucidated. Keywords Cryopreserved vein · Bypass stenosis · Neointimal hyperplasia · Angiogenesis · Chronic limb ischemia Abbreviations PAOD Peripheral arterial disease PTFE Polytetrafluoroethylene VSMC Vascular smooth muscle cell eNOS Endothelial nitric oxide synthase ECM Extracellular matrix VV Vasa vasorum VEGF Vascular endothelial growth factor DSA Digital subtraction angiography DAPI 4,6-Diamidin-2-phenylindol EC Endothelial cell CD Cluster of differentiation PECAM1 Platelet/endothelial cell adhesion molecule
Introduction Autologous vein is the favored conduit for bypass revascularization in chronic critical limb ischemia (CLI) regarding long-time patency and infection control (Conte 2010). If the saphenous or arm veins are absent or not eligible for surgery, prosthetic grafts made from polytetrafluoroetylene (PTFE) or polyester (Dacron®) are commonly used alternatives bearing inferior patency rates and the risk of graft infection. Allografts such as cryopreserved cadaveric veins are therefor considered a further alternative since 1976 (Harris et al. 2001).
13
Although the biological effects of cryoconservation on complex tissues such as a human vessel concerning reactivity, immunology and thrombogenicity are still elusive, the process itself has been well established and allografts are commercially available (Fahner et al. 2006; Muller-Schweinitzer et al. 2007). Feasibility in cardiac and peripheral bypass procedures as well as vascular access has been demonstrated, and clinical series report safe utilization in infectious and non-infectious sites (Faggioli and Ricotta 1994). One-year primary patency rates between 17 and 40 % for infrapopliteal reconstructions, however, emphasize clear indications regarding outcome, costs and frequency of subsequent interventions (Barshes et al. 2013; Farber et al. 2003; Harris et al. 2001; Martin et al. 1994; Walker et al. 1993). Arterialization of venous grafts is reflected by neointimal hyperplasia (NH), but may occur excessively or uncontrolled at the proximal anastomosis site, resulting in highly rigid lesions requiring aggressive treatment such as cutting balloon angioplasty or surgery enabling secondary patency rates of 45–63 % at 12 months (Kreienberg et al. 2007; Leseche et al. 1997; Zehr et al. 2011). Various stimuli concert NH. Endothelial denudation with exposure of procoagulant factors, platelet adhesion and leukocyte extravasation, vasospasm due to denervation and hypoxia, triggered by loss of vasa vasorum, mediates an acute injury response with consecutive extracellular matrix (ECM) degradation and monocyte homing (Collins et al. 2012; Davies and Hagen 2011). Antigenicity of allografts may further result in rapid initiation and progression of NH. Phenotype switch in vascular smooth muscle cell (VSMC), the main cells of the tunica media, from a quiescent contractile to a dedifferentiated proliferative type with migratory potential appears to be the main downstream effector (Lacolley et al. 2012). Additional loss of eNOSdriven endothelial relaxation response leads to a thickened, rigid neointima, predestined at sites with high turbulent flow and increased shear stress, such as the proximal anastomosis with proximity of host and graft tissue (Fig. 1; Collins et al. 2012; Muto et al. 2010). Angiogenesis within the tunica media is frequently seen in atherosclerotic plaque formation or venous grafts in coronary or peripheral position (Khurana et al. 2005; Westerband et al. 2000). Exact trigger mechanisms within the complex network of pro- and anti-angiogenic factors remain elusive. Expression of, i.e., VEGF due to tissue hypoxia might be of importance (Collins et al. 2012). Angiogenesis may as well be part of chronic vascular transplant rejection, but is, however, evident in NH. Various pharmaceutical and mechanical strategies have been assessed to improve outcome after autologous or allogeneic bypass procedures. Handling of the surgically harvested vein is a strong outcome determinant (Conte 2009;
13
Histochem Cell Biol
Santo et al. 2013). Preimplantation histological analysis has been proposed (Giannoukas et al. 1997). External sheathing and brachytherapy failed to prevent NH (Kohler et al. 1989). Promising results have come from topical drug application culminating in the PREVENT III trial, administering edifoligide, a VSMC modulator, which in the end did not improve outcome and therefor highlights the incomplete understanding of the pathways involved (Wiedemann et al. 2012). The cytostatic agents rapamycin and paclitaxel showed short-term inhibition but did not prevent NH after a total of 12 weeks (Murphy et al. 2007; Shukla and Jeremy 2012). The anti-angiogenic drug thalidomide attenuates experimental atherosclerosis but has never been used in intimal proliferation models. In summary, many strategies have been used to prevent NH, a phenomenon incompletely understood and supposedly in parts physiological, in a heterogeneous clinical setting. Thus, it seemed pertinent to further evaluate the morphological characteristics of NH in different human grafts. We present for the first time a detailed histological analysis of cryopreserved venous allograft stenosis due to NH in comparison with autologous bypass grafts 6 months after implantation.
Materials and methods Patients’ characteristics Stenosis samples were harvested during open surgical bypass repair 6 months after femorocrural revascularization with an ABO but not sex-matched cryopreserved human greater saphenous vein (CryoLife©) (specimen 1 and 2) or after an autologous greater saphenous vein femorocrural bypass (specimen 3 and 4), respectively. Control samples are of greater saphenous vein. Acquisition was in accordance with the Declaration of Helsinki, the local ethic review committee and with patients’ informed and written consent. Indications for revascularization procedures were CLI (patient 2) or tissue loss (patients 1, 3 and 4) with lack of autologous vein. Open bypass repair was due to recurrent rest pain and impaired lesion healing due to stenosis at the proximal anastomotic site seen on angiography (Fig. 1). All four patients had good arterial inflow at the time of operation and at least one patent crural target vessel with good pedal runoff. Bypasses were in reversed position without valvulotomy maneuver (see supplementary Table 1). Specimen acquisition and histochemistry Tissue samples were excised during operation and immediately fixated in 3, 5 % formaldehyde (Fischer,
Histochem Cell Biol
Fig. 1 Digital subtraction angiography image (a) and intraoperative situs photography (b) reveal a circumferential rigid stenosis (white arrow) with intimal thickening of the femorocrural cryopreserved vein allograft bypass of patient 2 at the proximal anastomosis at the femoral bifurcation (diamond), 6 months after implantation. The profunda femoris artery (asterisk) shows strong collateralization but only few calcifications. HE-stained sections of stenotic lesions of cryopreserved allografts (c, d patients 1, 2) and autologous vein grafts (f, g, patients 3, 4): all four stenotic specimen show a similar morphology distinct from the control specimen (e). A neointima (N) rich in extra-
cellular matrix is absent in the control vein. Here, the tunica media (M) consists of a collagen and elastin meshwork and vascular smooth muscle cells with a longitudinal portion luminal and a circular portion orientated toward the tunica adventitia (A) which has lost its orientation in the stenotic lesions (c, d, f, g). Small vessels (25 ± 11 µm diameter) spreading into the tunica media can be depicted in every graft sample but are completely absent in control samples. Pinkstained collagen is absent in the neointima of stenotic lesions as are elastic fibers. (Scale bar 100 µm, magnification ×40)
Table 1 List of all primary antibodies used in the study by target, identification number, manufacturer and dilution used in our experiments
Germany). One-micrometer samples were mounted on Superfrost© slides (Menzel, Germany) and primarily stained for hematoxylin and eosin (HE) and elastica van Gieson (EvG).
Target
Antibody
Manufacturer
Dilution
Ki67 CD31 CD34 Actin CD3 CD20 vWF CD68 Vimentin Desmin
ab16667 M0823 M7165 ab5694 ab16669 ab9475 F-3520 M0814 ab8978 ab32362
Abcam Dako Dako Abcam Abcam Abcam Sigma-Aldrich Dako Abcam Abcam
1:50 1:100 1:100 1:200 1:200 1:40 1:1000 1:100 1:200 1:100
Kiel Institute of Pathology
1:40,000
KiM1P
KiM1P is not commercially available, but has been used and described for many years in daily routine as well as experimental studies by the Institute for Pathology, University of Wuerzburg (further information is available upon request to the author)
Immunohistochemistry Table 1 shows the different antibodies used. Counterstaining was done with either nuclear fast red aluminum or Mayer’s hematoxylin (both from Carl Roth, Germany). Fluorescence Immunohistochemistry For double immunofluorescence, CD31 (ab28364, rabbit polyclonal, Abcam) and CD34 (M7165, mouse monoclonal, Dako) were conjugated with Cy3- and Cy5-labeled secondary antibodies (both from KPL, US). Counterstaining was DAPI (LifeTechnologies®).
13
Light microscopy Light microscopy images were made by Keyence BZ9000 microscope (Keyence, Japan). Image editing as well as size measurements within the samples was performed with AZ Analyser II software distributed by Keyence, Japan. Confocal laser scanning microscopy Confocal laser scanning microscopy images were taken using a Nikon Eclipse TI-E inverse confocal microscope equipped with Coherent sapphire lasers and an mbp communication visible fiber laser (lines: 405, 488, 561, 647 nm). Objective (Plan Apo 60x NA:1, 4) was used and detection in three simultaneous channels applied. The system was equipped with NIS-Elements Advanced Research Software (Nikon). ImageJ (1.47v, NIH, USA http://magej. nih.gov/ij) was used for editing of confocal laser scanning microscope pictures.
Results Microscopic and immunohistochemical examination of stenotic lesions revealed a uniform morphological pattern of autologous bypass and cryopreserved allograft tissue with distinct features compared with control vein specimen (Fig. 1). The neointima, lying between the endothelium and the lamina elastica interna, was between 142 and 760 µm thick (normal intima
