High pressure–low flow remodeling of the rat saphenous vein wall

Abstract

Objective

To better understand factors that may play a role in the development of varicosities.

Methods

We induced combined flow-pressure disturbance in the saphenous system of the rat by performing chronic partial clipping of the main branch. Biomechanical and quantitative histological testing was undertaken.

Results

A rich microvenous network developed. Bloodflow decreased to 0.65 ± 0.18 µl/s (control side, 3.5 ± 1.4 µl/s) and pressure elevated to 6.8 ± 0.7 mmHg (control side, 2.3 ± 0.2 mmHg, p < 0.05). Involution of the wall and lumen was observed (16.5%, 28.7% and 35.5% reduction in outer diameter, wall thickness and wall mass respectively, p < 0.05). Elevated macrophage (CD68) and cell division (Ki67) activity was observed. Elastic tissue and smooth muscle actin became less concentrated in the inner medial layers.

Conclusions

Low-flow induced morphological shrinking of the lumen in veins may override pressure-induced morphological distension. Loosening of the force-bearing elements during flow-induced wall remodeling may be an important pathological component in varicosity.

Keywords

Saphenous vein varicose veins microcirculatory changes venous disease

Introduction

There seems to be a consensus that one key factor in the development of subcutaneous venous varicosities is chronically elevated venous pressure in the lower extremities. Such arises due to gravitational effects determined by lifestyle and will be further exaggerated by valvular incompetence. Flow disturbances, inherited connective tissue weakness, obesity and local inflammation contribute to the pathomechanism.^1–5 Experimental elevation of venous pressure in the leg was achieved in our laboratory by keeping rats in tilted tube-cages for a chronic period, which doubled venous pressure⁶ and resulted in several massive cytophysiological alterations^7–10 in the venous wall. However, applied alone, it failed to induce varicous morphological remodeling.¹⁰ Meanwhile, a wider use of Duplex scan and high-resolution ultrasonographic devices in phlebological practice revealed a substantial flow disturbance in the majority of varicous networks.^11–13 Chronic flow alterations are known to induce morphological remodeling processes in the affected vessels.^14–17 Unfortunately, despite the emerging significance of flow disturbances in the development of pathological varicose morphology, no experimental study has yet addressed the flow-induced remodeling processes of the venous side of the circulation.

To investigate the effect of alterations in flow and pressure on the development of varicosities, we decided to induce flow disturbance in the saphenous system of the rat. Chronic partial occlusion of the main branch was performed and we examined the resulting biomechanical and quantitative histologic alterations of the main branch, caused by the combined effect of reduced flow and elevated pressure.

Methods

Partial chronic occlusion (“clipping”) of the saphenous vein

Experiments were performed on male Sprague–Dawley rats weighing 273 ± 40 (mean ± SD) g at the beginning of the experiments. All rats were anesthetized in a similar manner, 45 mg/kg bodyweight Nembutal i.p., repeated at reappearance of pain reflexes by one-third of this dose. During a sterile microsurgical operation, the most proximal segment of the main branch of the saphenous vein was dissected at a distance of 10 mm from its confluence with the deep femoral vein. A longitudinally slit 4 mm long semi-rigid, thick-walled piece of a plastic tube with a lumen diameter of 500 µm was placed around the vein (Figure 1(a)). This did not close the vessel but prevented any further increase in diameter as the animal grew.

Figure 1.

Main branch and collateral network morphology of the rat saphenous vein developed after eight week of partial occlusion. (a and b) In vivo video-microsopy of microprepared vein and network. Bars, 1 mm. 1, Plastic clip narrowing the main branch, 2, Main branch, 3, Collateral branch with retrograde flow. Arrows show direction of flow. Number marks are also valid for (c) to (f). (c) Monitor image of the venous segment (main branch) mounted for in vitro pressure angiography. Bar, 200 µm. (d) Batson 17 plastic cast, main branch with retrograde filling collateral branches. (e) Immunohistochemistry for smooth muscle actin (SMA) showing cross sections of main branch and collaterals, low magnification. (f) Video-microscopic picture of saphenous vein main branch with methylene-blue bolus moving upward. Horizontal line shows border of stain. See also retrograde filling side branches.

Demonstration of newly developed bypassing collaterals

After four and eight weeks, the animals were reanesthetized. In vivo videomicroscopic studies of the microsurgically dissected specimens were performed (six rats, Figure 1(a) and (b)) as well as Batson17 plastic casts were prepared through a popliteal side branch (12 rats, Figure 1(d)).

Pressure and flow in the strictured saphenous vein main branch

To test the effects of chronic partial occlusion on the hemodynamics, in a separate sets of rats (12 rats) venous pressure was measured in anesthetized animals through a microcannula (around 200 µm outer diameter) introduced into the saphenous vein main branch, distal to the level of the stricture, through a popliteal side branch. It was attached to a Braun pressure head and a low-pressure mainframe. Fifty microliter saline stained with methylene blue was injected into the vein and venous flow was determined by following the movement of the stained bolus on a long stretch of the saphenous vein by serial pictures (1 frame/s) by intravital microscopy (Figure 1(f)). Similar measurements were made on the control sides.

In vitro biomechanical testing of the main branch

Just below the clip, a side-branch free segment of the main branch was dissected and excised (12 + 8 rats, with four and eight weeks after chronic occlusion, respectively). The careful microsurgical preparation practically removed all loose connective tissue while leaving the dense connective tissue cover (adventitia) and smooth muscle (media) intact, as well as endothelial cells in place. The length of the removed (dissected) cylindrical segments varied between 8 and 12 mm. The identical segments of the contralateral side served as controls.

The cylindrical venous segments were mounted on 500 µm diameter glass cannulas, immersed in thermostated saline and stretched to their in vivo length in the glass-bottomed tissue bath of the pressure angiometer setup. Our angiometer was constructed of a pressure angiography tissue bath (Experimetria, Hungary), a Leica inverted microscope, a Leica DFC 320 digital camera, two servo-controlled pumps (Living Systems, Burlington, VT, USA) and an IBM PC with a monitor. The tissue bath was superfused with roller pumps (producing a flow of 2.5 ml/min). Intraluminal pressure was set by adjusting the servo-controlled pumps, pumping saline into the lumen. The image of the vessel segment appeared on the screen (Figure 1(c)), online. Off-line diameter measurements of the inner and outer diameter were made using the Leica QWin image analyzing software. Calibrations were made with a micrometer etalon (Wild, Heerbrugg, Switzerland). The normal-Krebs Ringer (nKR) solution was bubbled with a gas mixture of 95% oxygen and 5% carbon dioxide.

After a 30-min incubation period in normal Krebs–Ringer solution at 5 mmHg intraluminal pressure, pressure was increased in a stepwise manner, to 2–5–10–15–20 mmHg. Then at 5 mmHg intraluminal pressure, 10 µmol/l norepinephrine was added to the tissue bath and the measurement cycle was repeated to test contractility of the segments. With the norepinephrine still in the bath, 5.5 µmol/l acetylcholine was added to study endothelial dilation. Experiments were terminated by recording the characteristic pressure–diameter curves in Ca²⁺-free Krebs–Ringer solution (passive state). Biomechanical alterations of the chronically clipped saphenous vein segments were characterized by analyzing the pressure–radius characteristic curve. Spontaneous contraction (in nKR) and norepinephrine-induced contractions (10 µmol/l, inducing maximum contraction in rat saphenous vein segments), and ACh-induced relaxation (5.5 µmol/l, inducing maximum endothelial relaxation in rat saphenous vein segments) were compared to parameters measured at the same pressure levels in the relaxed state. Wall stress was computed using the Frank–Starling equation, σ = p*r_i/h, where p is the intraluminal (transluminal) pressure, r_i is the inner radius, and h is the wall thickness. Incremental elastic modulus was computed using the equation given by RH Cox,¹⁸ E_inc = (2r_o $r_{i}^{2}$ /( $r_{o}^{2}$ − $r_{i}^{2}$ )) × (Δp/Δr_o), where r_o and r_i are outer and inner radii, respectively, and Δr_o is the rise in outer radius in response to a pressure increase of Δp. The source of the drugs was Sigma (St Louis and Budapest).

Histology and immune-histology of the main branch

In a separate series, the same segments and surrounding tissue were removed for histological studies after four and eight weeks of clipping (13 animals), and analog specimens from contralateral sites served as controls. Histological samples were stained with hematoxylin-eosin (nuclei) and resorchin-fuchsin (elastica) and immune-histological staining was performed for CD68 (macrophage invasion), Ki67 (cellular division activity), and smooth muscle actin (SMA, contractile protein). The source of the antibodies was the R&D Sytems Inc. Secondary antibodies were visualized with the DAB technique. The Ventana Benchmark XT Immune-Automat System was used for the immune-histochemical staining. This ensured identical staining procedures and made the sections comparable with each other. For embedding the specimens, the tissue-array technique was used. Sections were scanned with the 3D Histech Pannoramic250 Scanner, and selected high magnification (pixel sizes 0.31 µm) RGB pictures (0–255 levels) were analyzed with the Leica QWin image analysis software. The green (for RF) and blue (DAB) levels were measured in a radial line using the “Measure profile” function, starting at the luminal surface and going radially in an outward direction. The colors (green for RF and blue for DAB with methylene blue background) were chosen after careful preliminary evaluation of color histograms.

Statistical analysis

Paired data (Control and Clipped side) were compared with paired t-tests. Pressure-radius, pressure-elasticity and pressure-contraction and relaxation curves were compared with two-way ANOVA tests. For quantitative histochemistry, five sections were selected for each segment, radial densities measured 3–5 times, and pooled data for the four groups were statistically analyzed. In specific cases, elastic modulus vs. RF density plots were constructed using the particular means of the separate groups.

Results

No swelling, edema or unapparent changes in gait were observed after four and eight weeks, respectively. Feeding habits and behavior of the animals remained unaltered also.

Bypassing collaterals

When incising the inguinal skin of reanesthetized animals, a rich collateral system appeared on the clipped side under video-microscopic examination. Brush-like conglomerates, consisting of several hundred small veins, running mostly parallel, were observed, and they were leading away from the affected main branch (Figure 1(a) and (b)). Some of them found their way toward collateral small branches bypassing the artificial stricture, and some of the latter dilated forming corrugated collaterals. We could prove that blood flow was leading away from the main branch via these newly developed veins by examining the direction they filled up, and by the serial pictures made after injecting either methylene blue or saline into one of the popliteal side-branches of the saphenous vein. It was possible to fill that collateral system from the direction of the saphenous vein main branch using Batson 17 plastic (Figure 1(d)). On the contralateral unclipped veins, such retrograde filling side-branches were scarcely observed, and existing valves in the vicinity of the confluence effectively closed retrograde flow. Cross sections of such multiple bypassing side-branches could also be observed on the histological sections (Figure 1(e)).

Hemodynamics of the main branch

As it was expected, chronic partial occlusion induced a substantial pressure rise in the more peripheral parts of the saphenous vein main branch. The venous pressure in comparison with the control side (anesthetized animals in the supine body position) doubled (Figure 2(a), p < 0.001 with two-way ANOVA). This was accompanied by a drastic drop in the main branch flow. After eight weeks of occlusion, control side flow of 3.5 ± 1.4 µl/s dropped to a mere 0.65 ± 0.18 µl/s at the side of the stricture (Figure 2(b), p < 0.001 with two-way ANOVA). Reduction of blood flow seems to be induced by diversion of blood from main branch toward the retrograde filling collateral system (Figure 1(a). and (b)).

Figure 2.

Hemodynamics and hemodynamically induced biomechanical remodeling of the saphenous vein main branch. Clipped sides compared with control sides. (a) Saphenous vein pressure. (b) Saphenous vein flow. (c) Relaxed diameter plotted against pressure in calcium-free medium. (d) Wall thickness at 10 mmHg in calcium-free medium. (e) Cross section of the wall. (f) Log elastic modulus plotted against intraluminal pressure. (g) Spontaneous tone as a function of pressure. (h) Norepinephrine-induced tone (10 µM/l) as a function of pressure. (i) Ach-induced (endothelial) dilation as a function of pressure. *p < 0.05, **p < 0.01, ***p < 0.001, significant difference between marked groups, according to one- and two-way ANOVA tests. ^#p < 0.05, ^##p < 0.01, different by paired t-test.

Biomechanical alterations of the main branch

Pressure–diameter curves clearly showed a reduction in diameter of the clipped segments in comparison with their contralateral unclipped controls (Figure 2(c)). At 10 mmHg intraluminal pressure for instance, the relaxed outer diameter of clipped segments reduced to 642 ± 29 µm in comparison with 764 ± 24 µm of the control side (p < 0.001 with the paired t-test). Corresponding values following eight weeks of clipping were 613 ± 30 and 734 ± 25 µm (p < 0.01 with paired t-test). Figure 2(d) and (e) shows that following eight weeks of clipping, the wall thickness values of the clipped segments are significantly less compared to those found in the contralateral side and this is the result of a reduced wall mass.

In Figure 2(f) we can observe that low-stress elastic modulus decreases between weeks 4 and 8 in controls, and such reduction is less at the clipped side. When elastic moduli were plotted against wall stress, after four weeks, the low stress modulus increased and the high stress modulus decreased in obliterated segments. Their values were at 0.5 kPa wall stress 4.36 ± 0.30 vs. 3.65 ± 0.22 and at 1.5 kPa wall stress 4.58 ± 0.15 vs. 4.88 ± 0.20, for clipped and control sides, respectively (logarithmic values in lgPa, statistically significant with ANOVA, p < 0.05). (Further analysis of elasticity-see later.)

The induced low flow–high pressure remodeling massively reduced contractility. While spontaneous tone increased between weeks 4 and 8 in control segments (p < 0.01), this was missing at the clipped side. As a result, after eight weeks, clipped segments exhibited much smaller spontaneous tone than control ones (Figure 2(g), p < 0.01). Similar observations could be made for the maximal contraction induced by norepinephrine (Figure 2(h)). The reduced contractility reached the level of statistical significance after four weeks of partial occlusion (p < 0.05). In addition, a reduced endothelial dilation capacity developed in venous segments after eight weeks of occlusion (Figure 2(i), p < 0.05).

Histological changes

Histological analysis proved the presence of both macrophage invasion (CD68, Figure 3(a) and (b)) and new cell formation (Ki67, Figure 3(c) and (d)) at the occlusion site. The presence of one or two cells with cell divison activity can be considered a normal feature (Figure 3(d)).

Figure 3.

Histological wall remodeling processes after four weeks of partial clipping. Immuno-histochemical stainings. (a) CD68 (macrophagic activity), clipped side; (b) CD68 control side; (c) Ki67 (cell division activity), clipped side; (d) Ki67, control side. Arrows show positively staining structures. Bars, 50 µm.

Resorchin-fuchsin (RF) staining revealed a massive remodeling of the elastic components in the wall. The inner elastic membrane is less condensed in clipped segments (Figure 4(a) to (d)). This is quantitatively demonstrated by the green color density measurements made in the radial direction outwardly from the luminal surface (Figure 4(e)). It did not show as low values in the vicinity of the endothelium of clipped side venous segments than in their unaffected contralateral controls (Figure 4(f) and (g), p < 0.01). This demonstrates a reduced compactness of the inner elastic membrane. Another observation is that peripheral elastic elements diminish between four and eight weeks in controls. This process seems to be less effective in clipped segments (see Figures 4(a) to (d) and the diagram of 4(f)). Thus, clipped and unclipped segments were statistically different (p < 0.01) not only around their minimums, but in the range of 0.3–10 µm (four weeks) and 3–17 µm (eight weeks). A comparison of RF densities with measured elastic parameters revealed a close to linear function between elastic densities in the range of 10–15 µm from luminal surface and elastic modulus measured in spontaneous tone at 10 mmHg (Figure 4(h)) or in the passive state at low pressure (Figure 4(i)) p < 0.01 for both).

Figure 4.

Remodeling of the elastic components after four and eight weeks of partial clipping. (a to e) Resorcin-fuchsin-stained (RF) sections. (a) Four weeks’ occlusion; (b) four weeks’ control; (c) eight weeks’ occlusion; (d) eight weeks’ control. Bar, 50 µm (valid for a–d). (e) Higher magnification insert of section ‘c’ showing a line of color density measurement and neighboring tissues. (f) Results of RF elastica density measurements in the green color as a function of distance from the luminal surface. Lower green values represent higher elastica density. Each curve represents 5–8 rats, and 15–24 measurements were made on each venous segment. (g) Minimum values of the above density curves showing maximum inner elastic membrane densities. Color code identical with that of (f). (h and i) Elastic moduli are plotted against RF green density in the range of 10–15 µm from the luminal surface. (h) Moduli of segments in the spontaneous tone at physiologically feasible higher pressure (10 mHg). (i) Passive moduli at low pressures (2 mmHg). Statistical differences with one- and two-way ANOVAs are shown, **p < 0.01 between clipped and controls at four weeks, ^##p < 0.01 between clipped and controls at eight weeks. ^††p < 0.01 statistical significance of the Pearson correlation (negative).

The applied low flow–high pressure hemodynamic challenge has also induced a remodeling of the contractile elements: age-induced accumulation of the contractile protein in the inner medial layers is less effective in clipped segments. At the same time, both at week 4 and 8, there is more contractile protein present scattered in the outer layers of the media in the occluded segments (Figure 5, p < 0.01 with two-way ANOVA).

Figure 5.

Remodeling of the contractile components after four and eight weeks of partial clipping. Smooth muscle actin (SMA-DAB) immune-histochemistry. (a) Four weeks’ occlusion; (b) four weeks’ control; (c) eight weeks’ occlusion; (d) eight weeks’ control. Bar, 50 µm (valid for a–d). (e) Density measured in blue color as a function of distance from the endothelial surface. **p < 0.01 between clipped and controls at four weeks, ^##p < 0.01 between clipped and controls at eight weeks.

Discussion

These studies unanimously confirmed that partial occlusion of the main branch of the saphenous vein did not lead to the expected varicosity-like pressure-induced dilation, but instead morphological involution of the wall and of the lumen was observed. This seems to be the result of chronic adaptation to reduced flow. There was a significant decrease of the passive outer diameter of partially occluded segments in comparison to controls at weeks 4 and 8 alike (Figure 2(c)). There is a characteristic elevation of wall thickness (Figure 2(d)) and wall mass (Figure 2(e)) of control side veins in the same period. This is missing at the partially occluded side inducing statistically significant differences between occluded and non-occluded sides. Histological analysis proved the presence of both macrophage invasion (CD68, Figure 3(a) and (b)), and new cell formation (Ki67, Figure 3(c) and (d)). We can conclude that acting chronically, the low–flow high pressure hemodynamic environment massively altered segmental geometry.

Several theories have been introduced to determine the relationship between pressure, flow and chronic venous disease.⁴ Mechanisms such as turbulence, shear force,¹ increased wall tension,¹⁹ and inflammatory cascades⁴ have been suggested to lead to changes in vessel wall structure, and alteration in constriction/relaxation properties seen in varicose veins. More recently, a theory was introduced naming impending venous drainage, and outflow obstruction as the underlying cause of the formation of varices.^20,21 An animal model developed in our laboratory was based on the commonly accepted theory that the formation of varicose veins is mainly driven by pressure loading. Rats were submitted to chronic gravitational loading and several adaptive and pathological alterations were identified. However, classic pathological remodeling characteristic for chronic venous disease with the appearance of reticular veins, teleangiectasias and varicose dilations, torques or plaques was not observed.^6–10

As a further step to reveal the role of hemodynamics in the remodeling of the venous wall, we created a model, where flow disturbance induces the elevated venous pressure in the saphenous vein. In our recent model of chronic stricture on the proximal segment of the saphenous vein in rats, a rich collateral system appeared on the tributary venous system of the clipped side. Our hemodynamical measurements showed a reduced flow and doubling of venous pressure in the strictured main branch (Figure 2(a) and (b)). Venous inflow into the main branch was directed toward the newly developed bypassing venous routes with “inverted flow” flow (Figure 1). After a combined biomechanical–histological study of the main branch, we can conclude that the wall had sufficient strength to withhold doubling of luminal pressure without any morphological distension. On the contrary, a morphological reduction of the lumen was observed, which, with all probability was the result of a flow induced wall remodeling process (Figure 2(c)).

Such massive changes in wall geometry cannot be expected to occur without parallel changes in wall elasticity. Figures 2(c) and (4) demonstrate that really that is the case. As measured by quantitative histochemistry, the structure of the inner elastic membrane of the wall of the occluded segments loosened and became less dense (Figure 4) which corresponds with reduced passive elastic modulus at the physiologically relevant 1.5 kPa level wall stress. Elastic elements outward from the inner elastic membrane were also affected. A spontaneous reduction of such elements between weeks 4 and 8 is less effective. Our analysis suggests that the presence of fragmented dense elastic elements in the outward vicinity of the inner elastic membrane, at a distance of 10–15 µm from the endothelial surface (Figure 4(e)), is an important factor determining elastic modulus at low pressures in the passive state (Figure 4(i)) and also at higher pressure with some smooth muscle tone (Figure 4(h)).

Another important observation was reduced contractility. Both spontaneous tone²² and norepinephrine induced tone decreased (Figure 2(g) and (h)). Actin density in the inner medial layers did not differ between clipped and non-clipped sides (Figure 5), and this cannot explain the observed significant reduction in the contractile force. The presence of more scattered smooth muscle elements in the outer layers of the venous wall, however, marks that contractile elements are also affected by the low flow–high pressure remodeling process. There is an accumulation of the contractile proteins in the inner medial layers between weeks 4 and 8 even in controls with substantial elevation of contractile force. This process is less effective in clipped segments, or results in accumulation of scattered contractile elements in the outer layers without any improvement of contractility.

The almost total loss of endothelial dilation capacity after eight weeks of clipping (Figure 2(h)) cannot occur without pathological consequences. According to analogies observed on resistance arteries, it can be connected with lowered flow and be responsible for lumen shrinkage.

Observations on how vein structure, geometry, elasticity, and contractility are affected by varicose disease are divergent. Evidence suggests that the formation of varicose veins may be due to the weakness of the vein wall as a result of significant increase in intimal and medial thickness and elevation of collagen content in the media with concomitant decrease of elastin content.²³ The reduced amount and fragmented state of elastic fibers,²⁴ dysregulation of the synthesis of collagen I and III in smooth muscle cells,²⁵ and overexpressed aFGF in the wall via FGFR and the MAP kinase pathway may influence expression of enzymes involved in extracellular matrix metabolism²⁶ and play a role in vein wall remodeling, leading to the described rigidity.²⁴ It has been suggested that smooth muscle cells derived from varicose veins are less differentiated and demonstrate increased proliferative and synthetic capacity than those derived from normal veins and may therefore contribute to the weakening capacity of the vein wall to resist pressure.²⁷ An ultrasound study suggested that varicose vein sections may show dilatation of the lumen compared with the control with no change in the wall thickness itself. Wall thickness compensation in this case is by production of collagen instead of smooth muscle.²⁸

In conclusion, partial obstruction of the saphenous vein in rats resulted in the development of a rich bypassing microvenous network reminiscent of that seen in lower limb varicosity. Contrary to expectations, the main branch did not dilate morphologically, instead involution with morphologically reduced lumen, reduced wall thickness and reduced wall mass was observed. We are convinced that this type of remodeling occurred as low flow induced remodeling (morphological shrinkage), and has overridden the passive distention induced by chronically elevated venous pressures. We succeeded to demonstrate that elastic modulus is dependent on the presence and amount of fragmented elastic elements in the inner media, but outward from the inner elastic membrane. The scattering of contractile elements was accompanied by a substantial loss of contractile force. In case of both elastic and contractile elements, not only the amount of tissue, but also their distribution in the different layers of the media has a decisive mechanical role.

Such observations, we believe, should be taken into consideration when explaining the pathomechanism of varicose disease.

Footnotes

Acknowledgements

We would like to thank Oraveczné Murányi Ildikó for her work as a laboratory assistant. The authors would also like to thank Dr Zoltan Varady, Venenklinik, Frankfurt am Main, Germany for his valuable advice and continuous support. The expert technical assistance of Oraveczné Murányi Ildikó is also appreciated.

Ethical approval

The experiments have been approved by the Semmelweis University’s local Animal Ethics Committee and by state authorities (permission numbers 22.1/2960/003/2009 and PEI/001/801-2/2015).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The paper was supported by the Hungarian National Grants OTKA TO32019, TO42670, the Hungarian Kidney Foundation, and by a grant from the Dean of the Medical Faculty of the Semmelweis University.

References

Fegan

Kline

. The cause of varicosity in superficial veins of the lower limbs. Brit J Surg 1972; 59: 798–801.

Brand

Dannenberg

Abbott

et al.

The epidemiology of varicose veins: the Framingham Study. Am J Prev Med 1988; 4: 96–101.

Lee

Evans

Allan

et al.

Lifestyle factors and the risk of varicose veins: Edinburgh Vein Study. J Clin Epidemiol 2003; 56: 171–179.

Bergan

Pascarella

Schmid-Schönbein

et al.

Pathogenesis of primary chronic venous disease: insights from animal models of venous hypertension. J Vasc Surg 2008; 47: 183–192.

Gemmati

Federici

Catozzi

et al.

DNA-array of gene variants in venous leg ulcers: detection of prognostic indicators. J Vasc Surg 2009; 50: 1444–1451.

Monos

Contney

Cowley

Jr. et al.

Effect of long-term tilt on mechanical and electrical properties of rat saphenous vein. Am J Physiol 1989; 256: H1185– H1191.

Monos

Bérczi

Nádasy

. Local control of veins: Biomechanical, metabolic, and humoral aspects. Physiol Rev 1995; 75: 611–666.

Monos

Raffai

Dornyei

et al.

Structural and functional responses of extremity veins to long-term gravitational loading or unloading-lessons from animal systems. Acta Astronaut 2007; 60: 406–414.

Raffai

Lodi

Illyes

et al.

Increased diameter and enhanced myogenic response of saphenous vein induced by two-week experimental orthostasis are reversible. Physiol Res 2008; 57: 175–183.

10.

Lorant

Nadasy

Raffai

et al.

Remodeling of the rat saphenous vein network in response to long-term gravitational load. Physiol Res 2003; 52: 525–531.

11.

Delis

. Leg perforator vein incompetence: functional anatomy. Radiology 2005; 235: 327–334.

12.

Ibegbuna

Delis

Nicolaides

. Haemodynamic and clinical impact of superficial, deep and perforator vein incompetence. Eur J Vasc Endovasc Surg 2006; 31: 535–541.

13.

Labropoulos

Tassiopoulos

Bhatti

et al.

Development of reflux in the perforator veins in limbs with primary venous disease. J Vasc Surg 2006; 43: 558–562.

14.

Carmeliet

. Angiogenesis in life, disease and medicine. Nature 2005; 438: 932–936.

15.

Dumont

Loufrani

Henrion

. Key role of the NO-pathway and matrix metalloprotease-9 in high blood flow-induced remodeling of rat resistance arteries. Arterioscler Thromb Vasc Biol 2007; 27: 317–324.

16.

Osol

Mandala

. Maternal uterine vascular remodeling during pregnancy. Physiology (Bethesda) 2009; 24: 58–71.

17.

Baeyens

Schwartz

. Biomechanics of vascular mechanosensation and remodeling. Mol Biol Cell 2016; 27: 7–11.

18.

Cox

. Three-dimensional mechanics of arterial segments in vitro: methods. J Appl Physiol 1974; 36: 381–384.

19.

Raffetto

Khalil

. Mechanisms of varicose vein formation: valve dysfunction and wall dilation. Phlebology 2008; 23: 85–98.

20.

Gaweesh

. Impeded venous drainage: novel view of chronic venous disease pathophysiology. Med Hypotheses 2009; 73: 548–552.

21.

Neglen

Thrasher

Raju

. Venous outflow obstruction: An underestimated contributor to chronic venous disease. J Vasc Surg 2003; 38: 879–885.

22.

Berczi

Greene

Dornyei

et al.

Venous myogenic tone: studies in human and canine vessels. Am J Physiol 1992; 263: H315–H320.

23.

Elsharawy

Naim

Abdelmaguid

et al.

Role of saphenous vein wall in the pathogenesis of primary varicose veins. Interact Cardiovasc Thorac Surg 2007; 6: 219–224.

24.

Kirsch

Wahl

Bottger

et al.

Primary varicose veins-changes in the venous wall and elastic behavior. Chirurg 2000; 71: 300–305.

25.

Sansilvestri-Morel

Rupin

Badier-Commander

et al.

Imbalance in the synthesis of collagen type I and collagen type III in smooth muscle cells derived from human varicose veins. J Vasc Res 2001; 38: 560–568.

26.

Kowalewski

Malkowski

Sobolewski

et al.

Evaluation of aFGF/bFGF and FGF signaling pathway in the wall of varicose veins. J Surg Res 2009; 155: 165–172.

27.

Xiao

Huang

Yin

et al.

In vitro differences between smooth muscle cells derived from varicose veins and normal veins. J Vasc Surg 2009; 50: 1149–1154.

28.

Mahmoud

Refaat

. Smooth muscle changes in varicose veins: an ultrastructural study. J Smooth Muscle Res 2001; 37: 123–135.