Journal of Vascular Surgery
Volume 41, Issue 2 , Pages 303-311, February 2005

An animal model of venous hypertension: The role of inflammation in venous valve failure

Presented at the16th Annual Meeting of the American Venous Forum, Orlando, Fla, Feb 29, 2004.

  • Luigi Pascarella, MD

      Affiliations

    • Department of Surgery, University of California, San Diego
    • ,
  • Geert W. Schmid-Schönbein, PhD

      Affiliations

    • Department of Bioengineering, University of California, San Diego
    • ,
  • John Bergan, MD (FACS, FRCS Hon (Eng))

      Affiliations

    • Department of Surgery, University of California, San Diego
    • Corresponding Author InformationReprint requests: John Bergan, MD, 9850 Genesee, Suite 410, La Jolla, Ca 92037

Received 10 August 2004; accepted 19 October 2004.

Article Outline

Background

Clinical observation suggests that chronic venous insufficiency is related to failure of venous valves. Duplex ultrasound studies of lower extremity superficial veins regularly show valve failure and venous reflux. Gross morphologic observation of venous valves in surgical specimens shows tearing, splitting, scarring, and disappearance of valves.

Hypothesis

Venous valve damage is acquired, linked with venous hypertension, and affected by inflammation.

Objective

The objective of this study was to investigate the inflammatory process in valve remodeling associated with acute and chronic venous hypertension.

Methods

A femoral arteriovenous fistula was created in study animals (Wistar rats, n = 60), and animals without an arteriovenous fistula were studied as controls (n = 5). At 1, 7, 21, and 42 days animals with the femoral arteriovenous fistula were anesthetized, and systemic pressure, the pressure in the femoral vein distal to fistula, and the pressure of the femoral vein in the contralateral hind limb were measured. Timed collection of blood backflow after division of the femoral vein distal to the fistula and in the alive, anesthetized animal was collected, measured, and calculated per unit time to be used as an indicator of valve insufficiency. The femoral vein distal to the fistula was harvested; valvular structures were examined and measured. Specimens were processed, and longitudinal sections were made and challenged with immunostaining antibodies against matrix metalloprotease (MMP)-2 and MMP-9. Sections were examined, and expression of molecular markers was determined by light absorption measurements after image digitization.

Results

One week after the procedure, all animals exhibited some degree of hind limb edema ipsilateral to the arteriovenous fistula. Pressure in the femoral vein distal to the fistula was markedly increased on average to 96 ± 9 mm Hg. Reflux was increased in a time-dependent manner, with the 21-day and 42-day groups showing the highest values. Valves just distal to the fistula showed an increased diameter of the valvular annulus and a shortening of the annular height. Venous wall findings included fibrosis and fusion of the media and adventitia and scarring and disappearance of valves principally in the 21- and 42-day specimens. Immunolabeling for MMP-2 showed an increased level in the 21- and 42-day groups. MMP-9 showed an increased level at 1 day, followed by a more marked level in the 21- and 42-day groups.

Conclusions

In this animal model of venous hypertension the findings of limb edema, increasing valvular reflux, and morphologic changes of increased annulus diameter and valve height are seen. Histologic changes included massive fibrosis of media and fusion with adventitia. Inflammatory markers MMP-2 and MMP-9 are strongly represented, and valve disappearance occurs after these markers are present. The gross morphologic changes seen are quite similar to those observed in human surgical specimens removed in treatment of venous insufficiency.

Clinical relevance

When observed angioscopically at the time of vein stripping, saphenous vein valves show severe deformities including shortening, scarring, and tearing. The current model of induced venous hypertension demonstrates early venous valve changes that replicate those observed in humans. This observation provides a link from venous hypertension to an induced inflammatory reaction that stimulates the valve damage. Thus the model could be useful for defining the fundamental mechanisms that cause venous valve failure and varicose veins and in pharmacologic testing to prevent or treat venous insufficiency.

 

Clinical observation suggests that chronic venous insufficiency is related to failure of superficial and deep venous valves. Although there is a large body of literature addressing deep venous incompetence, it is the gross morphologic changes in superficial valves that have attracted careful study. Duplex ultrasound studies of lower extremity superficial veins regularly show valve failure and venous reflux. 1 Careful angioscopic preoperative study of 116 patients by Yamaki et al 2 showed 28% with elongated and atrophic valve cusps, 48% with expanded and depressed valve commissures, and even 33% with absent valves. 2

In a morphologic study of 65 superficial venous valves Corcos et al 3 found changes of both thickening and thinning. Thickening of the distal portion of the cusp was related to the collagen component that shortened and crumpled the valve. Thinning in the commissural regions led to dilation (aneurysm formation). Forty years earlier Cotton 4 found dilation of the vein wall projecting beneath the cornua and widening the cornual angle. He said, “… of 156 valves which were examined, 123 were macroscopically normal; only 33 were sclerosed, The sclerosis was often so marked that only the rim of the valve was left at its attached border.”

Proximal veins in the lower extremities are subject to pressure fluctuations produced by intra-abdominal compression and abdominal wall stresses. Kistner 5 has suggested that primary valve incompetence develops from stress factors such as straining and coughing. Surgical specimens of saphenous veins with reflux confirmed by duplex ultrasound scanning have more extensive leukocyte infiltration and membrane adhesion molecules more numerous on proximal surfaces of venous valves and vein walls than on distal surfaces. 6

Observations of histologic sections from varicose veins harvested during therapeutic surgical procedures done for chronic venous insufficiency and study of venous blood samples from patients with chronic venous insufficiency suggest that in venous disease significant tissue inflammation occurs, and multiple indicators of cell activation can be found. 7 Inflammatory indicators include attachment of leukocytes to the endothelium, infiltration of monocytes, lymphocytes, and mast cells into connective tissue, and development of fibrotic tissue infiltrates. 8

This study was done to investigate the inflammatory process associated with acute and chronic venous hypertension. The working hypothesis was that venous valve damage is acquired, linked with venous hypertension, and affected by the inflammatory process associated with acute and chronic venous hypertension. Although the model requires arterialization of a vein segment, there is no evidence that aterializing a vein segment produces an inflammatory reaction or venous valve damage.

Back to Article Outline

Materials and methods 

The animal procedures in this study have already been reviewed and approved by the Animal Subjects Committee of the University of California San Diego. The animals were kept in an approved animal facility in accordance with University of California San Diego policies, provisions set forth by National Institutes of Health, and all federal, state, and local laws and regulations governing the use of animals in research. Postprocedural record keeping policies were followed according to policies established by the Animal Subjects Committee and Office of Campus Veterinary Services at University of California San Diego.

The animal model was the Wistar rat (250 to 300 g, 8-week-old male; Charles River Laboratory, Inc, Wilmington, Mass). The study consisted of two groups of animals: (1) animals with the femoral arteriovenous anastomosis (n = 15 killed at each time point) and (2) control group (n = 5) and the contralateral femoral vein.

Previous studies showed a gradual and a time-dependent disappearance of the venous valve in the presence of the fistula. Therefore we studied four time points for the animals with the femoral arteriovenous fistula: 1 day, 7 days, 21 days, and 42 days.

Surgical procedure. Each animal was given general anesthesia (pentobarbital sodium, 50 mg/kg, intramuscular) with supplemental doses of 5 mg/kg when needed. Systemic blood pressure, cardiac activity, and respiratory functions were monitored during the entire procedure. Leg motion and toe reflexes were checked for evaluation of the anesthetic level.

To prevent blood coagulation, heparin was administered (1000 U/kg body weight) before clamping both the artery and the vein. A femoral arteriovenous fistula was created in the groins of 60 animals. The arteriotomy length was 0.5 mm, and the anastomosis was constructed with 10-0 monofilament suture.

Ligation of the femoral vein above the fistula and ligation of the superficial epigastric vein were also performed to increase the blood pressure in the superficial and deep venous systems of the limb and to reduce the risk of cardiac failure (Fig 1). Cardiac failure in the original study described in the doctoral thesis of van Bemmelen, University of Amsterdam, 1984, was 25%.

At the end of each time point the animals with the femoral arteriovenous fistula were given general anesthesia. The systemic pressure, the pressure in the femoral vein 3 cm distal to fistula, and the pressure of the femoral vein in the contralateral hind limb were measured. The pressure measurements were made by catheterization of the contralateral femoral vein and contralateral artery and catheterization of the popliteal vein ipsilateral to the fistula. No other pressures were measured.

The blood backflow through the valve after division of the femoral vein distal to the fistula and in the alive, anesthetized animal was collected, measured, and calculated per unit time to be used as a measure of valve insufficiency.

The femoral vein distal to the fistula was harvested, and although a longitudinal section, the valvular structures were examined and measured. The diameter of the annulus and the valvular height were measured (Fig 2). The femoral vein was then fixed in 4% paraformaldehyde for 24 hours.

Processing of each specimen included the dehydration in graded ethanols and the embedding in paraffin. Longitudinal sections (6 μm of thickness) were made, mounted on poly- (L-lysine) coated slides, and challenged with the immunostaining. Primary antibodies against MMP-2 (sc-10736; Santa Cruz Biotechnology Inc, Santa Cruz, Calif), and MMP-9 (sc-6840; Santa Cruz Biotechnology Inc), were applied. The ABC method was used to detect the expression of the previously mentioned enzymes by using techniques described elsewhere. The ABC method is an immunoperoxidase procedure for localizing antigens. 9 Secondary biotinylated antibodies are applied with peroxidase-conjugate-avidin to yield a dark purple color for MMP-2 and a dark-brown reaction product for MMP-9. The sections were examined with light microscopy at several magnifications up to ×100 objective (numeric aperture 1.4). In addition to direct microscopic observations, the expression of molecular markers was determined by direct light absorption measurements after image digitization on a laboratory computer (Image J-NIMH, Bethesda, Md) and expressed in terms of grade 0 (unstained), grade 1 (slightly stained), grade 2 (moderately stained), and grade 3 (strongly stained).

Statistics. Data were expressed in terms of means and standard deviation. An analysis of variance test was applied. Values of P <.01 and <.05 were considered significant.

Back to Article Outline

Results 

One week after the procedure, all animals exhibited some degree of hind limb edema ipsilateral to the arteriovenous fistula (Fig 3).

Hemodynamic measurements. With a mean systemic pressure of 126 ± 5 mm Hg, the mean pressure in the femoral vein distal to the fistula was markedly increased to 96 ± 9 mm Hg (P < .05), compared to pressures in the controls and in the contralateral femoral vein (mean pressures, 9 ± 9 mm Hg).

Timed collection of the blood backflow (reflux) in the presence of the native femoral pressure was increased in a time-dependent manner, with the 21-day and 42-day groups showing the highest values compared to the controls (P < .01 for the 21- and 42-day values and P < .05 for the 7 days compared with the controls) (Fig 4). The presence of an increased backflow reflects increasing valvular incompetence and increased reflux flow.

  • View full-size image.
  • Fig 4. 

    This graph shows progressively increasing reflux through the first valve distal to the arteriovenous fistula (P < .01 for the 21- and 42-day values and P < .05 for the 7-day group compared with the controls).

Valve morphology. Macroscopic observation of the terminal valve in the saphenous vein and of the valve just distal to the fistula in the femoral vein showed an increased diameter of the valvular annulus and a shortening of the annular height (Fig 5, A and B). This was seen in the animals with the arteriovenous fistula but not in the controls. These changes between operated limbs versus controls developed in a time-dependent manner later (P < .01 at 21 and 42 days from the creation of the fistula).

  • View full-size image.
  • Fig 5. 

    A, This graph illustrates progressive valve dilation caused by the high pressure early and venous wall weakness later (P < .01 at 21 and 42 days from the creation of the fistula). B, Progressive valve shortening is shown in this graph (P < .01 for the 21- and 42-day values compared with the controls).

Bulging of the valve and dilation of the commissures were observed in the 1- and 7-day groups. Later, stretching, shortening of the valvular leaflets is seen (Fig 5), and ultimately the complete disappearance of the valves was observed in the 21- and 42-day groups.

Histology. Microscopic observations were made in transverse and longitudinal sections of the saphenous and the femoral veins. These showed abnormalities of the venous wall and of the valves.

The venous wall showed disappearance of the media accompanied by a massive venous wall fibrosis. Thickening of the venous wall was observed principally at 21 and 42 days from creation of the fistula (Fig 6). The valve showed lesions including the stretching and a shortening of the leaflets mentioned above progressing to complete disappearance of valvular structures in the 42-day group (Fig 6, B).

  • View full-size image.
  • Fig 6. 

    A, At 21 days loss of tissue architecture is prominent. There is fibrosis of the media and fusion with the adventitia. The valve is present, and strong staining for MMP-2 is present in the media. (Transverse sections, mono-staining for MMP-2 with Nova-red background staining). B, At 42 days the valve has disappeared. There is still fibrosis of the media and fusion with the adventitia. MMP-2 staining is prominent. (Transverse sections, mono-staining for MMP-2 and Nova-red background staining.)

Immunohistochemistry. Staining of MMP-2 and MMP-9 was carried out on longitudinal sections of the valves (Fig 7), and the venous wall and optical density measurements were calculated (NIH Image software).

  • View full-size image.
  • Fig 7. 

    These figures are longitudinal sections of vein wall and valve leaflet double stained for MMP-2 and MMP-9 with Nova-red background staining. The control vein without arteriovenous fistula is on the left. 21-day specimen in the center shows MMP-9 staining of the intima, MMP-2 staining of the media, and the valve is present. The 42-day specimen (right) shows similar staining and absence of the valve.

The immunostaining for MMP-2 showed an increased expression in the 21- and 42-day groups (Fig 8) compared to the 1- and 7-day groups and to the controls (P < .01). MMP-2 was expressed mostly in the subintima and the media of the venous wall and in the subendothelial connective tissue of the valvular leaflets (Fig 7).

The immunostaining for MMP-9 showed an initial increased peak expression at 1 day compared with controls and the other groups (P < 0.01) and a second, more marked peak expression in the 21- and 42-day groups (P < .01). Thus the expression of MMP-9 seems to be bimodal with two expression peaks, one at 1 day after the surgery and the second at 21 and 42 days from the surgery. MMP-9 was mostly expressed in the intima and subintima of both the venous wall and valvular leaflets (Fig 9).

Back to Article Outline

Discussion 

This study shows that a gradual remodeling of venous valves develops when subjected to venous hypertension. Immediately after placement of the arteriovenous fistula, a passive viscoelastic dilation of the annulus occurs that is accompanied by shortening of the valve leaflet with reduction of the leaflet overlap. It is followed by more gradual changes of the vein that progress during the first week and cause a time-dependent increase in reflux. Between 7 and 21 days a marked increase in reflux occurs, and this progresses through the 42nd day. During this interval, progressive inflammation is seen as indicated by the MMP-2 and MMP-9 levels. The MMPs serve to break down extracellular collagen. Finally, stunting, scarring, and shortening of venous valves are encountered that mimic the gross morphologic changes that are observed in saphenous veins that are removed surgically. 6, 8

But venous pressure does not act as an isolated phenomenon. It alters fluid shear stress and affects numerous endothelial cell, leukocyte, and platelet functions. 7 It selectively up-regulates intercellular adhesion molecule–1 expression 10, 11 and other manifestations of the inflammatory response with leukocyte rolling, adhesion, and migration. 12 These responses to acute venous hypertension have also been observed in the model of acute mesenteric venous hypertension and reperfusion.

Although the model requires arterialization of a vein segment, there is no evidence that arterializing a vein segment produces an inflammatory reaction or venous valve damage. In the present preparation, pulsations of the arteriovenous fistula are transmitted to the venous wall. The pressures recorded are in the range observed in veins at the ankle in the lower extremities of humans. Cyclic strain on the venous endothelial cells activates extracellular signal-regulated kinases 1 and 2. 13 It also shifts the redox potential and enhances oxidative stress. 14, 15

The pathogenesis of venous insufficiency remains unclear. It is believed by some that valve failure leads to superficial reflux and dilation of distal veins. 16 There might also be a reduction of the mechanical strength of the vein wall. This facilitates dilation of the venous annulus, separation of the valve cusps, and subsequent reflux that increases with venous hypertension.

The present study supports both possibilities. Valve remodeling is part of the inflammatory reaction induced by venous hypertension. The hypertension alters shear stress and incites the leukocyte-endothelial interactions that initiate rolling, firm adhesion, and migration of the leukocytes. Macrophages accumulate in the extraendothelial tissue and might synthesize MMPs, which in turn compromise the strength of the vein wall. On the one hand, free radical production and the proteolytic action of the MMPs and, on the other hand, new synthesis of collagen by growth factors derived from macrophages might then cause the histologic alterations observed in the vein wall. 14 The current studies served to identify the MMP levels. Their activity remains to be identified with zymographic techniques.

In studying varicose vein sections, Wali et al 17, 18 observed an increase in collagen, which disrupted the normal palisade arrangement of the intima and the sheet-like arrangement of medial smooth muscle cells. In the present preparation of induced hypertension, profound fibrosis of the vein wall was seen in the 42-day specimens. Similarly, Kirsch et al 19 observed increased fibrosis of the extracellular matrix of varicose veins. Much of this was due to an increase in collagen IV. In discussing the histomorphologic changes of the vein wall at the saphenofemoral junction, Stücker et al 20 stated that the alterations observed were “…regarded by us as a reaction to altered perfusion and pressure in the veins.”

Back to Article Outline

Conclusions 

In this animal model of venous hypertension the findings of tissue edema, valvular reflux, and morphologic changes with increased annulus diameter and valve height were seen.

Histologic changes included massive fibrosis of media and fusion with adventitia.

Enhanced levels of MMP-2 and MMP-9 were present, and valve leaflets disappeared after these markers became evident.

The gross morphologic changes seen were quite similar to those observed in human surgical specimens removed in treatment of chronic venous insufficiency. Although the model requires arterialization of a vein segment, there is no evidence that aterializing a vein segment produces an inflammatory reaction or venous valve damage.

Back to Article Outline

References 

  1. Myers KA , Wood SR , Lee V , et al.   Variations of connections to the saphenous system in limbs with primary varicose veins (a study in 1481 limbs by duplex ultrasound scanning) . J Phlebology . 2002;2:11–17
  2. Yamaki T , Sasaki K , Nozaki M . Preoperative duplex-derived parameters and angioscopic evidence of valvular incompetence associated with superficial venous insufficiency . J Endovasc Ther . 2002;9:229–233
  3. Corcos L , De Anna D , Dini M , Macchi C , Ferrari PA , Dini S . Proximal long saphenous vein valves in primary venous insufficiency . J Mal Vasculair . 2000;25:27–36
  4. Cotton LT . Varicose veins (gross anatomy and development) . Br J Surg . 1961;48:589–598
  5. Kistner RL . Primary venous valve incompetence of the leg . Am J Surg . 1980;140:218–244
  6. Ono T , Bergan JJ , Schmid-Schönbein GW , Takase S . Monocyte infiltration into venous valves . J Vasc Surg . 1998;27:158–166
  7. Fukuda S , Schmid-Schönbein GW . Fluid stress in leukocytes . In:  Schmid-Schönbein GW ,  Granger N editor. Molecular basis for microcirculatory disorders . New York: Springer; 2003;p. 169
  8. Takase S , Lerond L , Bergan JJ , Schmid-Schönbein GW . Enhancement of reperfusion injury by elevation of microvascular pressures . Am J Physiol Heart Circ . 2002;282:H1387–H1394
  9. Hsu SM , Raine L , Fanger H . The use of autiavidin antibody and avidin-biotin-peroxidase technics . Am J Clin Path . 1981;75:816–821
  10. Nagel T , Resnick N , Atkinson WJ , Dewey CF , Gimbrone MA . Shear stress selectively upregulates intercellular adhesion molecule-1 expression in cultured human vascular endothelial cells . J Clin Invest . 1994;94:885–891
  11. Simon SI , Goldsmith HL . Leucocyte adhesion dynamics in shear flow . Ann Biomed Eng . 2002;30:315–332
  12. Schmid-Schönbein GW , Takase S , Bergan JJ . New advances in the understanding of the pathophysiology of chronic venous insufficiency . Angiology . 2001;52(Suppl 1):S27–S34
  13. Ikeda M , Takei T , Mills I , Kito H , Sumpio BE . Extracellular signal-regulated kinases 1 and 2 activation in endothelial cells exposed to cyclic strain . Am J Physiol Heart Circ Physiol . 1999;H614–H622
  14. Cheng JJ , Wung BS , Chao YJ , Hsieh HJ , Wang DL . Cyclic strain induces redox changes in endothelial cells . Chin J Physiol . 1999;42:103–111
  15. Awolesi MA , Widmann MD , Sessa WC , Sumpio BE . Cyclic strain increases endothelial nitric oxide synthase activity . Surgery . 1994;116:439–444
  16. Goldman MP , Fronek A . Anatomy and pathophysiology of varicose veins . Dermatol Surg . 1989;15:138–145
  17. Wali MA , Dewan M , Eid RA . Histopathologic changes in the wall of varicose veins . Int Angio . 2003;22:188–193
  18. Wali MA , Eid RA . Changes of elastic and collagen fibers in varicose veins . Int Angio . 2002;21:337–343
  19. Kirsch D, Schreiber J, Dienes H-P, Bòttger T, Juninger T. Alterations in the cellular matrix of venous walls in varicose veins. Vasa;28:95-9.
  20. Stücker M , Krey T , Schulz-Ehrenburg U , Altmeyer P . The histomorphologic changes at the saphenofemoral junction in varicosis of the greater saphenous vein . Vasa . 2000;29:41–46

 Competition of interest: none.

PII: S0741-5214(04)01370-9

doi:10.1016/j.jvs.2004.10.038

Journal of Vascular Surgery
Volume 41, Issue 2 , Pages 303-311, February 2005

Article Tools

* Image Usage & Resolution