European Journal of Vascular & Endovascular Surgery
Volume 40, Issue 1 , Pages 117-121, July 2010

Oestradiol Levels in Varicose Vein Blood of Patients with and without Pelvic Vein Incompetence (PVI): Diagnostic Implications

Department of Vascular Surgery, St. Josef Hospital, Bochum, Germany

Received 14 June 2009; accepted 10 January 2010. published online 04 March 2010.

Article Outline

Abstract 

Purpose

To assess the difference in the oestradiol levels of blood taken from varicose veins in patients with and without pelvic vein incompetence (PVI).

Materials and methods

Women of child-bearing age with symptomatic primary or recurrent varicose veins of the great saphenous vein (GSV) were included in a prospective study. Patients underwent duplex ultrasonography and pelvic vein phlebography. They were divided into a group with PVI (PVI group) and a control group with GSV reflux alone (VV group). Blood samples were collected from the GSV at the sapheno-femoral junction or lower in the thigh as well as from the arm. Oestradiol levels were determined by electroluminescence.

Results

Between January and December 2007, 40 women were studied, of which 19 showed phlebographic evidence of PVI (PVI group), while 21 were included in the VV group. Phlebography revealed an incompetent ovarian vein in 14 (74%) patients of the PVI group, dilated uterine and ovarian plexuses in 12 (63%) and an incompetent internal iliac vein in six cases (32%). In the PVI group, the median oestradiol level in GSV samples was 121pgml−1 (range: 12–4300), while in the VV group the median level was 75pgml−1 (range: 9–1177). In the upper limb, the PVI group patients had a median level of 78pgml−1 (range: 15–121) and the VV group patients 68pgml−1 (range: 13–568). The ratio of lower limb/upper extremity was significantly higher (p<0.002) in patients of PVI group (median: 1.9; range: 0.7–33) than in those of the VV group (median: 1.1; range: 0.8–13). A threshold ratio of 1.4 showed the highest combined sensitivity and specificity in differentiating patients with PVI from those without.

Conclusions

In patients with varicose veins arising from the GSV, oestradiol levels were significantly higher in the lower limb than in the upper extremity in the subgroup with associated PVI. It may be possible to use this observation as a diagnostic test in patients with suspected PVI. This deserves further study.

Keywords: Pelvic vein incompetence, Oestradiol levels, Limb varicosity, Phlebography

 

The term pelvic vein incompetence (PVI) describes pathological reflux in pelvic veins. It is often associated with pelvic pain, dyspareunia and pelvic heaviness and might also play a role in the development of varicose veins.1, 2 Invasive and non-invasive diagnostic methods have been used to detect pelvic reflux. The detection of reflux at the sapheno-femoral junction (SFJ) originating in the pelvis on ultrasonography (US) might lead to the suspicion of PVI.3 However, only diagnostic methods requiring the use of contrast media guarantee reliable detection of pelvic reflux pathways.4, 5, 6 Female sex hormones are mostly produced in the ovary. Incompetence of pelvic veins could allow reflux of hormone-rich blood to the SFJ and reach the great saphenous vein (GSV).

Oswald et al.7 have already measured the hormone content of varicose veins to assess the role of ovarian vein incompetence as a source of varicose veins. Building on this experience, we assumed that if high sex hormone levels could be detected in the lower limb in patients with phlebographic signs of PVI, this could help to select patients who should undergo contrast media examinations, reducing diagnostic costs and increasing patient acceptance.

We therefore conducted a prospective study analysing oestradiol blood levels in the lower and upper limbs in female patients with primary or recurrent GSV varicose veins and phlebographic evidence of PVI. We compared these findings to a control group of patients with isolated GSV varices.

Back to Article Outline

Materials and methods 

Female patients of child-bearing age with symptomatic primary or recurrent GSV varicose veins were recruited between January and December 2007 at our vascular centre. The study protocol was approved by our local ethics committee. Patients were divided into two subgroups: a PVI and a VV group, depending on the presence of PVI on phlebography. Not all patients underwent phlebography, which was only undertaken in those where clinical or duplex ultrasound scanning suggested that this might be present. The decision to perform phlebography of the pelvic veins was based on clinical signs typical of PVI: a history of vulval varicosities during pregnancy, an increase in venous congestion symptoms related to the menstrual period, dyspareunia, lower abdominal pain of unknown origin as well as a history of repeated surgery for recurrent varicose veins at the SFJ. PVI was also suspected if varicose veins were found at atypical sites (buttock, perineal or vulval region). Duplex ultrasound examinations were performed in all patients and if reflux was detected in the groin originating from cranial (epigastric) and/or medial (pubic and pudendal) veins, PVI was also suspected and pelvic venography was performed regardless of the above-mentioned symptoms. All patients gave written informed consent for their inclusion in this study.

Exclusion criteria included age under 18 years; known vascular malformation in the abdomen or lower limb; pregnancy; thrombophlebitis of the GSV; a history of deep vein thrombosis (DVT) in the lower limb; and previous allergic reactions to radiological contrast medium. Patients who met the inclusion criteria underwent a meticulous clinical examination. Previous surgery due to lower limb varicosity, pregnancies as well as clinical extension of varicosities (based on Clinical–Etiology–Anatomy–Pathophysiology (CEAP) classification8) were recorded.

Phlebography was performed using an imaging technique based on the descriptions of Ahlberg.9 The examination proceeded in a standardised fashion with visualisation of the ovarian veins (OVs) and the common iliac veins (CIVs) in an antero-posterior projection and of the internal iliac veins (IIVs) in an oblique projection (30° left or right anterior oblique).

The presence and the extent of any reflux were documented and the names of the pelvic veins affected by reflux were recorded. PVI was considered to be present if one of the following criteria was fulfilled:

-pelvic reflux toward the ipsi- or contralateral proximal thigh;

-visualisation of a dilated OV throughout its entire course;

-retrograde filling of the main stem of the IIV and at least one tributary (gluteal, ischiadic or obturator veins); or

-retrograde filling of contrast medium across the midline.

Based on our previous experience,10 an OV was considered to be dilated if its diameter was at least 1.5 times that of the contralateral vessel. If the vein was dilated without reflux, it was not considered incompetent. In selected cases, the main ovarian trunk and/or the IIV were embolised with the use of standard spring coils (Gianturco spring embolisation coils; Cook). Complications were of investigation and treatments were recorded.

Blood samples were collected after venous duplex ultrasound imaging and diagnosis of GSV incompetence leading to varicose veins. The GSV was marked with the help of US at the groin or the proximal part of the medial thigh. Blood samples were collected as near to the SFJ as possible using a 23-gauge needle with the patients in upright position and performing a Valsalva manoeuvre. In the upper limb, blood was taken from an ante-cubital vein.

The levels of the active form 17-beta-oestradiol were determined by sandwich electrochemiluminescent immunoassay (ECLIA) using the Elecsys 2010 system (Roche, Mannheim, Germany – coefficient of variation 2–8%). The time of blood sampling with reference to the menstrual cycle, hormone replacement therapy or oral contraception was considered irrelevant due to the intention to compare oestradiol levels in the upper and lower limbs and not to analyse absolute values.

Statistical analysis 

To compare the study and the control groups, we used the Mann–Whitney U-test for ordinal data (assuming non-normal distribution of data in view of the wide scatter of results) and Fisher's exact test categorical data. Descriptors for ordinal data in this study are the median, interquartile range and range of data. Regarding the comparison of oestradiol levels, in patients with bilateral varicose veins, the higher value measured in the lower limb was considered for analysis. The maximum difference between lower limb measurements accepted as reliable was 5%. A difference of less than 20% between varicose vein blood and upper limb blood oestradiol levels was considered non-significant.

In order to define the most predictive lower limb/upper extremity ratio for the diagnosis of PVI, we calculated specificity and sensitivity for a range of values of this ration and plotted a receiver operating characteristic (ROC) curve.

Back to Article Outline

Results 

Between January and December 2007, 40 female patients were included in this study, of which 19 (48%) patients had clinical signs and symptoms as well as US findings of PVI and were included in PVI group, while 21 (52%) patients were assigned to the VV group. Demographic data are shown in Table 1. Amongst all patients, there were 13 cases of varicose vein recurrence after previous stripping of the GSV and in 27 cases of primary varicose veins (see Table 1: demographic data). The mean number of pregnancies was 2.5 per patient (range: 0–5). Thirty (75%) patients presented with CEAP clinical class C2 and 10 (25%) with class C1. There was no case of oedema of the legs or venous ulceration. Varicose veins in the thigh and the calf were the main clinical findings. In the PVI group, 10 patients (53%) had varicose tributaries originating in the gluteal or groin region and connecting to varicose veins on the thigh or on the calf.

Table 1. Demographic data.
PVI groupVV groupp-Value
Age mean (range)42 (29–55)36 (18–52)0.043 (+)
Primary varicosity9 (47%)4 (19%)0.091 (×)
Recurrence10 (53)17 (81%)

CEAP
Class C17 (37%)3 (14%)0.15 (×)
Class C212 (63%)18 (86%)

(+) Mann–Whitney-U-test, (×) Fisher's exact test.

The standardised phlebography protocol was completed at the time of hospitalisation in all 19 patients in the PVI group and confirmed PVI in all cases. There were no adverse events due to the contrast agent nor was there any case of bleeding or pelvic vein thrombosis. An incompetent OV was recognised in 14 (74%) patients of the PVI group, dilated plexuses in 12 (63%) and an incompetent IIV in six cases (32%). Endovascular treatment with coil obliteration of the incompetent veins was performed in 10 cases.

The oestradiol measurements are summarised in Table 2, including the lower limb/upper limb ratios. These data have been plotted in Fig. 1a and b, confirming that the lower limb/upper limb ratios are higher in the PVI group. In order to define the most predictive ratio for diagnosis of PVI, we plotted the ROC curve (Fig. 2). A threshold ratio of 1.4 showed the highest combined sensitivity and specificity in differentiating patients with PVI from those without.

Table 2. Oestradiol levels (pgml−1). In patient with bilateral varicosities, the higher oestradiol level in the lower limbs was considered for comparison. See text for explanation. All data are median (range).
PVI groupVV groupp-Value (+)
Lower limb121 (12–1177)75 (9–4300)N.S.
Upper extremity68 (13–568)68 (15–121)N.S.
Lower limb/upper extremity ratio1.1 (0.7–33)1.9 (0.8–13)0.002
Modified lower limb/upper extremity ratio1.1 (0.7–1.8)1.9 (0.8–4.7)0.001

(+) Mann–Whitney-U-test.

  • View full-size image.
  • Figure 1 

    Oestradiol levels in the study groups. Depicted is the ratio lower limb/upper extremity. Fig. 1a includes all patients, while in Fig. 1b patients with the highest levels of both groups were excluded (see text for explanation). Box and whisker plots show median (central line), upper and lower quartiles (box) and range excluding outliers (whiskers).

  • View full-size image.
  • Figure 2 

    ROC curve. Sensitivity and specificity of ratios are reported. The distance of the curve from the middle values is correlated to the predictive value of the ratio as tool to differentiate patients with PVI from patients without PVI. A threshold ratio of 1.4 showed the highest combined sensitivity and specificity in differentiating patients with PVI from those without.

Back to Article Outline

Discussion 

In 1947, Taylor11 used the term ‘venous congestion syndrome of the pelvis’ to describe a combination of symptoms consisting of pelvic varicosities, dysmenorrhoea, post-coital pain and pelvic pain. This classical tetrad of symptoms is associated with a typical pattern of varicosities.1, 2 In some cases, incompetent pelvic veins may carry reflux to the venous system of the lower limb, resulting in primary varicosities and/or in recurrence after surgical treatment of varicose veins.2, 12 Symptoms often worsen during menstruation,13 the higher prevalence in multiparous women3 as well as the positive therapeutic effects of hormonal replacement on symptoms14, 15, 16 suggest that sex hormones might play a crucial role in the pathophysiology of this peculiar clinical entity.

High oestradiol levels can result from an excessive medical treatment,17 at the beginning of the menopause18 and rarely due to oestradiol-producing tumours.19 By contrast, lower oestradiol levels can be caused by an endocrine disease.20 In addition, hormonal regulation is influenced by psychiatric diseases, chronic stress, extreme sport or anorexia.21, 22, 23 Oestradiol is transported through the ovarian veins to the left renal vein as well as to the inferior vena cava. After liver inactivation, it is eliminated through the kidneys. In patients with incompetent pelvic or ovarian veins, the oestradiol-rich blood may flow to the lower limb veins through multiple connections between the pelvic and the lower limb venous systems.

In our study population, hormone levels were higher in the lower limb than in the upper limb only in patients where PVI was associated with varicose veins. The patients with PVI were slightly older than the VV group patients. These data confirm the observations reported by Oswald et al.,7 who assessed hormone levels of varicose veins to detect the incompetence of the OV as a source of varices. These authors measured the concentration of hormones such as oestrogen, corticoids and oestradiol in varicose vein blood of 10 patients with pre-menstrual pain and 10 without. In this group of patients, there was no significant difference between the hormone concentration levels in any of the samples in the group without pain. In the symptomatic group, there was no difference between the oestrogen and corticoid blood concentration. However, in the symptomatic group, the concentration level of oestradiol in the sample from painful varicose veins was two to nine times higher compared with samples from the non-affected leg and brachial vein. Oswald speculated that the higher concentration of oestradiol was due to an incompetent ovarian vein, which follows a route through the pelvis to the varicose veins and which may explain the high oestrogen content. The severity of PVI was not be classified in our study so we were unable to analyse the correlation between the severity of PVI and oestradiol levels. We cannot speculate as to whether higher oestradiol levels may be helpful in detection of PVI without phlebography.

We observed a wide range of oestradiol levels but it is likely that this may have arisen in part because of the different menstrual phase in which blood samples were collected. Failing to record this aspect should be considered a drawback of this study. Patients were not asked about medical intake of hormones and/or oral contraceptives, which would also influence absolute hormones levels. Unrecognised endocrine diseases may also have played a role. We aimed to minimise these differences by basing our conclusions on hormone ratios rather than absolute hormone levels. Other limitations of our study include omission of transvaginal ultrasound imaging. Due to a lack of experience with transvaginal US of the OV, we decided not to rely on this diagnostic tool. Some authors have reported that combined trans-abdominal and transvaginal US is potentially useful for non-invasive screening to determine which patients with chronic pelvic pain may benefit from selective ovarian venography.24 We avoided undertaking phlebography in patients without signs and symptoms of PVI so we cannot be sure that all patients in the VV group were free of pelvic vein incompetence. The consequences of exposing women of child-bearing age to radiation and the already-reported high sensitivity and specificity of US3 in detecting PVI were the main reasons for avoiding the use of phlebography in asymptomatic patients. It is also unclear whether there is any value in treating asymptomatic pelvic vein reflux. Finally, our study involved relatively few patients and a much larger cohort would have been required to provide more exact data in view of the variability of oestradiol levels.

Back to Article Outline

Conclusions 

These preliminary data indicate that blood samples taken from the GSV in the lower limb could help to identify those patients with clinical signs and symptoms of PVI who should undergo phlebography. If our findings can be confirmed in larger cohorts of patients, it may be possible to minimise the costs of diagnosis for patients with PVI, avoiding exposure to radiation and contrast media.

Back to Article Outline

Conflict of interest 

None.

Back to Article Outline

Funding 

None.

Back to Article Outline

References 

  1. Scultetus AH, Villavicencio JL, Gillespie DL, Kao TC, Rich NM. The pelvic venous syndromes: analysis of our experience with 57 patients. J Vasc Surg. 2002;36:881–888
  2. Perrin MR, Labropoulos N, Leon LR. Presentation of the patient with recurrent varices after surgery (REVAS). J Vasc Surg. 2006;43:327–334discussion 334
  3. Geier B, Barbera L, Mumme A, Köster O, Marpe B, Kaminsky C, et al. Reflux patterns in the ovarian and hypogastric veins in patients with varicose veins and signs of pelvic venous incompetence. Chir Ital. 2007;59:481–488
  4. Laissy JP, Cinqualbre A, Loshkajian A, Henry-Feugeas MC, Crestani B, Riquelme C, et al. Assessment of deep venous thrombosis in the lower limbs and pelvis: MR venography versus duplex Doppler sonography. Am J Roentgenol. 1996;167:971–975
  5. Greiner M, Gilling-Smith GL. Leg varices originating from the pelvis: diagnosis and treatment. Vascular. 2007;15:70–78
  6. Hobbs JT. Varicose veins arising from the pelvis due to ovarian vein incompetence. Int J Clin Pract. 2005;59:1195–1203
  7. Oswald M, Heintz C, Wolhlaib U. How to prove incompetence of the ovarian vein as a source of varicogenesis by testing the hormone content of the varicose veins of the leg. In: Phlebolymphology. Special issue for European Congress of UIP, Bremen; 1999.
  8. Eklöf B, Rutherford RB, Bergan JJ, Carpentier PH, Gloviczki P, Kistner RL, et al. Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg. 2004;40:1248–1252
  9. Ahlberg NE, Bartley O, Chidekel N, Edlundh KO. Roentgenological diagnosis of pelvic varicosities in women. Acta Obstet Gynecol Scand. 1965;43(Suppl. 1):
  10. Asciutto G, Mumme A, Marpe B, Köster O, Asciutto KC, Geier B. MR venography in the detection of pelvic venous congestion. Eur J Vasc Endovasc Surg. 2008;36:491–496
  11. Taylor HC. Pelvic pain based on a vascular and autonomic nervous system disorder. Am J Obstet Gynecol. 1954;67:1177–1196
  12. Labropoulos N, Leon L, Engelhorn CA, Amaral SI, Rodriguez H, Kang SS, et al. Sapheno-femoral junction reflux in patients with a normal saphenous trunk. Eur J Vasc Endovasc Surg. 2004;28:595–599
  13. Ciardullo AV, Panico S, Bellati C, Rubba P, Rinaldi S, Iannuzzi A, et al. High endogenous oestradiol is associated with increased venous distensibility and clinical evidence of varicose veins in menopausal women. J Vasc Surg. 2000;32:544–549
  14. Farquhar CM, Rogers V, Franks S, Pearce S, Wadsworth J, Beard RW. A randomized controlled trial of medroxyprogesterone acetate and psychotherapy for the treatment of pelvic congestion. Br J Obstet Gynaecol. 1989;96:1153–1162
  15. Boivin P, Cornu-Thenard A, Charpak Y. Pregnancy-induced changes in lower extremity superficial veins: an ultrasound scan study. J Vasc Surg. 2000;32:570–574
  16. Grandas OH, Mountain DH, Kirkpatrick SS, Cassada DC, Stevens SL, Freeman MB, et al. Regulation of vascular smooth muscle cell expression and function of matrix metalloproteinases is mediated by estrogen and progesterone exposure. J Vasc Surg. 2009;49:185–191
  17. Rolnick SJ, Kopher RA, DeFor TA, Kelley ME. Hormone use and patient concerns after the findings of the Women's Health Initiative. Menopause. 2005;12:399–404
  18. Sowers MR, Randolph J, Jannausch M, Lasley B, Jackson E, McConnell D. Levels of sex steroid and cardiovascular disease measures in premenopausal and hormone-treated women at midlife: implications for the “timing hypothesis”. Arch Intern Med. 2008;168:2146–2153
  19. Ghazi AA, Mofid D, Rahimi F, Marandi H, Nasri H, Afghah S. Oestrogen and cortisol producing adrenal tumour. Arch Dis Child. 1994;71:358–359
  20. Mueller SO, Korach KS. Estrogen receptors and endocrine diseases: lessons from estrogen receptor knockout mice. Curr Opin Pharmacol. 2001;1:613–619
  21. Baker FC, Driver HS. Circadian rhythms, sleep, and the menstrual cycle. Sleep Med. 2007;8:613–622
  22. Jacoangeli F, Masala S, Staar Mezzasalma F, Fiori R, Martinetti A, Ficoneri C, et al. Amenorrhea after weight recover in anorexia nervosa: role of body composition and endocrine abnormalities. Eat Weight Disord. 2006;11:e20–e26
  23. Finstad SE, Emaus A, Potischman N, Barrett E, Furberg AS, Ellison PT, et al. Influence of birth weight and adult body composition on 17beta-oestradiol levels in young women. Cancer Causes Control. 2009;20:233–242
  24. Park SJ, Lim JW, Ko YT, Lee DH, Yoon Y, Oh JH, et al. Diagnosis of pelvic congestion syndrome using transabdominal and transvaginal sonography. Am J Roentgenol. 2004;182:683–688

PII: S1078-5884(10)00059-6

doi:10.1016/j.ejvs.2010.01.023

European Journal of Vascular & Endovascular Surgery
Volume 40, Issue 1 , Pages 117-121, July 2010

Article Tools

* Image Usage & Resolution