Width of the Great Saphenous Vein Lumen in the Groin and Occurence of Significant Reflux in the Sapheno-femoral Junction

Aim: This study was performed to assess the contribution of the width of the anechogenous lumen of the great saphenous vein (GSV) in the groin measured by ultrasound (US) to the diagnostics of haemodynamically signifi cant refl ux (HSR) in the sapheno-femoral junction (SFJ). Methods: We examined 200 lower limbs with primary varicose veins in 182 patients. Duplex scanning was performed with the patients in the supine position. Longitudinaly imaging (B-mode) the inner anechogenous diameter of GSV was measured 4–5 cm distal to SFJ. PW Doppler sampling volume was placed at the same distance. The refl ux was elicited by Valsalve manoeuvre. The HSR was defi ned as a backfl ow lasting ≥ 1 s with a velocity of Vmax ≥ 10 cm/s.


INTRODUCTION
In ultrasound examination for venous insuffi ciency of the lower extremities we use duplex scanning to assess the function of venous valves and search for the refl ux.Morphological evaluation of the vein in B-mode (when we focus on the measurement of the anechogenous venous lumen) is only secondary.In case of venous dilatation we consider venous insuffi ciency.Sometimes it is possible to scan venous valves, their motility and reaction to Valsalva manoeuvre or manual compression.However, minor venous valves are diffi cult to scan.The new generation of ultrasonographic devices makes it possible to follow their function and blood-fl ow under both physiological and pathological conditions 12 .In lower extremities with primary varicose veins we can observe insuffi ciency of the sapheno-femoral junction (SFJ) most frequently 7,15 .The cause of trunk varices of the great saphenous vein (GSV) is refl ux in the terminal and preterminal valves in the sapheno-femoral junction 3 .Recent work has already evaluated the relation between the width of GSV and the occurrence of the refl ux in SFJ, and found a statistically signifi cant correlation between the competence of the terminal valve and a GSV diameter of < 5 mm in the proximal third of the thigh (p < 0.001).When the GSV diameter was 6 mm, an incompetent terminal valve was more likely (p < 0.001) (ref. 4).
The aim of this validation study was to objectively assess the usefulness of ultrasound scanning of the width of the GSV lumen in the diagnostics of refl ux in SFJ.Specifi cally the objective was to assess the signifi cance of ultrasound scanning of the width of the anechogenous GSV lumen in the groin in B-mode in the diagnostics of haemodynamically signifi cant refl ux in SFJ.

MATERIALS AND METHODS
We examined 200 lower extremities with primary varicose veins in 182 patients (125 women and 57 men, mean age 52±12 years).The extremities were examined using duplex sonography for clinical manifestation of venous insuffi ciency of C0 -C6, according to the CEAP classifi -cation.Duplex sonography (linear 10 MHz probe, Logiq 5 Pro, General Electric Co.) was performed with patients in the supine position 6,9 .
In B-mode using longitudinal imaging, the measured inner anechogenous diameter of the GSV was 4-5 cm distally from SFJ and PW Doppler (Pulsed Wave Doppler) sample was placed at the same distance.The refl ux was elicited by Valsalva manoeuvre presented for 5-10 seconds, and was defi ned as a backfl ow lasting ≥ 1 s (ref. 1,6,10,11 ).Haemodynamically signifi cant refl ux had a backfl ow velocity of Vmax ≥ 10 cm/s.
The examined great saphenous veins were subclassifi ed according to the presence of haemodynamically signifi cant refl ux in SFJ into two groups: GSV with proven haemodynamically signifi cant refl ux (Vmax ≥ 10 cm/s) and the GSV without haemodynamically signifi cant refl ux.In both groups we established maximum and minimum width of lumen and calculated means, medians and standard deviation (SD).The groups were compared using a t-test.
To objectively evaluate how useful measuring the width of the anechogenous lumen of the GSV is for haemodynamically signifi cant refl ux in SFJ, we used a contingency table (Table 2).The boundary width of anechogenous lumen of GSV was 5 mm.Then we calculated sensitivity, specifi city, positive predictive value, negative predictive value and diagnostic accuracy of the measurements of width of anechogenous lumen for hameodynamically signifi cant refl ux in the SFJ.

RESULTS
Haemodynamically significant ref lux (Vmax ≥ 10 cm/s) was found in 152 GSVs with a mean width of anechogenous lumen of 6.39 mm (minimum 2.7 mm, maximum 13.0 mm), median 6.0 mm, SD ± 2.21 mm.No haemodynamically signifi cant refl ux was found in 48 GSVs.In the latter group, the mean width of anechogenous lumen was 4.41 mm (minimum 2.5 mm, maximum 6.8 mm), median 4.4 mm, SD ± 0.96 mm (Table 1).The diff erence between upper quartiles of width of GSV lumens with haemodynamically signifi cant refl ux and without refl ux was statistically signifi cant (p < 0.01); the diff erence between lower quartiles was statistically signifi cant (p < 0.05).The diff erence in the mean width of the GSV of both groups was also statistically signifi cant (p < 0.01) (Table 1, Chart 1).
Sensitivity of a GSV dilation of ≥ 5 mm for the presence of haemodynamically signifi cant refl ux in the SFJ was 69.7 %.Specifi city of the GSV dilation of ≥ 5 mm for the presence of hameodynamically signifi cant refl ux in the SFJ was 64.6 %.Positive predictive value of the GSV dilation of ≥ 5 mm for the presence of haemodynamically signifi cant refl ux in the SFJ was 86.2 %, negative predictive value was 40.3 %, and diagnostic accuracy of the GSV dilation of ≥ 5 mm for the presence of hameodynamically signifi cant refl ux in the SFJ was 68.5 %.

DISCUSSION
Primary varicose veins of lower extremities are characterized by progressive worsening of valve function and dilation of the venous lumen.The most frequent fi nding in non-complicated (stage C1-3 according to CEAP) and complicated (stage C4-6 according to CEAP) varicose veins is SFJ and GSV trunk insuffi ciency -68 % and 83 %, respectively, of lower the extremities 14 .SFJ incompetence predominates in primary as well as in recurrent varices 17 .
According to our clinical experience and the literature, the normal width of the GSV below the groin does not exceed 5-6 mm (ref. 8).Varicose veins have, in comparison with normal veins, a greater diameter of the lumen and hypertrophy of venous wall, particularly intima.This is due to increased collagenous fi bre 16 .In an extensive prospective study including 612 lower extremities with GSV trunk varices, the mean width of the GSV with refl ux in the thigh (88 %) was 5.75 mm, the mean width of the GSV without refl ux (12 %) was 4.98 mm (ref. 15).In our work the boundary value of the width of anechogenous GSV lumen (to assess SFJ insuffi ciency) was 5 mm.The mean width of insuffi cient GSV in our set (n=152) was greater than reported by Sadouni et al., i.e. 6.39 mm (SD ±2.21 mm) (Table 1).
PW Doppler sample location was distal to the termination of the saphenous junctional tributaries, below the preterminal valve, in order to assess the function of the SFJ.The preterminal valve is located 3-5 cm below the terminal valve, distal to the termination of junctional tributaries to the GSV, to prevent refl ux from the veins when the terminal valve is closed 3,4 .
The reflux in SFJ was assessed in the course of Valsalva manoeuvre according to the maximum velocity of the backfl ow (Vmax) measured by PW Doppler 4-5 cm below the termination of the GSV to the common femoral vein because the velocity of refl ux and the volume of the regurgitating venous blood corresponded both to the clinical grade of venous insuffi ciency and the scope of the refl ux.In contrast, the refl ux duration does not correspond to the clinical grade of chronic venous insuffi ciency according to the CEAP (ref. 5), nor to the scope of refl ux 13 .
In our set of 200 GSV the diff erence between the upper and lower quartiles of lumen width of GSVs with haemodynamically signifi cant refl ux and those without refl ux was statistically signifi cant (p<0.01 resp.p<0.05) (Table 1).The diff erence in mean width of the GSVs of both groups was also statistically signifi cant (p<0.01).However, the sensitivity of dilation of the GSV ≥ 5 mm below the groin due to haemodynamically signifi cant re-Width of the great saphenous vein lumen in the groin and occurence of signifi cant refl ux in the sapheno-femoral junction fl ux in the SFJ was not high (69.7 %).Measurement of the width of anechogenoue lumen of GSV in B-mode does not recognize about 30 % of lower extremities with signifi cant refl ux in the SFJ.This implies that about one third of the GSVs, with a width below the groin not exceeding 5 mm -in ultrasonic scanning in B-mode, is aff ected by haemodynamically signifi cant refl ux in SFJ, confi rmed by PW Doppler.Dilation of the GSV ≥ 5 mm below the groin was less specifi c (64.6 %).Therefore, between the GSVs with anechogenous lumen < 5 mm we will fi nd about 35 % with signifi cant refl ux in the SFJ.
The positive predictive value of the width of anechogenous lumen GSV ≥ 5 mm below the groin was relatively high (86.2%).If GSV ≥ 5 mm is confi rmed, there is a probability of 86.2 % that the duplex scanning proves haemodynamically signifi cant refl ux in the SFJ.In contrast, the negative predictive value of dilation of the GSV ≥ 5 mm was very low (40.3 %).Hence, if a dilation of the GSV ≥ 5 mm below the groin in B-mode is not found, then there is only 40.3 % probability that the SFJ is functional and the duplex scanning will not prove haemodynamically signifi cant refl ux.The diagnostic accuracy of dilation of the GSV ≥ 5 mm is 68.5 %.This means that in approx.70 % of the GSV ≥ 5 mm haemodynamically signifi cant refl ux in SFJ is present.
The diff erence between GSVs without refl ux and those with haemodynamically signifi cant refl ux was statistically signifi cant (p < 0.05, and p < 0.01) for maximum, minimum and mean width of anechogenous lumen (Table 1).However, the results show that for the diagnosis of hameodynamically signifi cant refl ux in the SFJ, measurement of the width of the anechogenous lumen of the GSV be-low the groin is insuffi ciently sensitive (69.7 %), as well as insuffi ciently specifi c (64.6 %).Only 68.5 % of the all measurements of width of the GSV below the groin in B-mode provided accurate indirect assessment of the functions of valves in the SFJ.
In a patient the fi nding of dilated the GSV ≥ 5 mm below the groin means a 86.2 % probability that the duplex (p < 0.01) scanning will confi rm haemodynamically signifi cant refl ux in SFJ.The fi nding of non-dilated the GSV < 5 mm below the groin shows only 40.3 % probability that the SFJ is functional and without haemodynamically signifi cant refl ux.Therefore, ultrasound scanning of the width of anechogenous lumen of the GSV below the groin (B-mode) may serve only as an orientation examination to assess the function of valves in SFJ.Only the duplex scanning (PW Doppler) can answer the question whether there is haemodynamically signifi cant insuffi ciency of the SFJ.

Table 1 .
Statistical comparison of widths of anechogenous lumens of the GSV without refl ux and the GSV with haemodynamically signifi cant refl ux in the SFJ

Table 2 .
Haemodynamically signifi cant refl ux in the SFJ (Vmax.≥ 10 cm/s) and the width of the anechogenous GSV lumen in the groin