Comparative study between foam sclerotherapy and sandwich technique for management of ovarian vein reflux.

The object of this work was to compare and emphasize the outcome of two endovenous techniques for management of ovarian vein reflux, including the foam sclerotherapy technique (group A) and the sandwich technique (group B).

Conception and design: MM, IA Analysis and interpretation: MM, TI Data collection: MM, MS Writing the article: MM, MS Critical revision of the article: MM, TI, IA Final approval of the article: MM, MS, TI, IA Statistical analysis: Not applicable Obtained funding: Not applicable Overall responsibility: MS

None.

This prospective randomized controlled study was carried out on 40 female patients with PCS that lasted more than 6 months and did not respond to medical therapy, with refluxing ovarian vein and a diameter larger than 6 mm. The study was conducted from March 2024 to December 2025 after being approved by the Ethical Committee of Kafr Elsheikh University, Kafr Elsheikh, Egypt. Informed written consent was obtained from the patients. Patients were divided into two treatment groups (group A, foam sclerotherapy; and group B, sandwich technique) by using a computer-generated randomization sequence with allocation concealed in sealed opaque envelopes. Each group was further stratified, based on ovarian vein diameter, into 6- to 8-mm and >8-mm subgroups. Exclusion criteria were female patients with CPP due to gynecologic diseases (endometriosis, adenomyosis, leiomyoma, and pelvic inflammatory disease); pelvic congestion secondary to May–Thurner syndrome or Nutcracker syndrome or internal iliac vein reflux; presenting with CPP due to urologic, gastrointestinal, or musculoskeletal diseases; allergy to iodinated contrast media; chronic kidney disease; and pregnancy. All patients were subjected to full history taking, clinical examination, urine analysis, pregnancy test, laboratory examination (complete blood count, prothrombin time, international normalized ratio, partial thromboplastin time, and kidney function tests), and obstetric consultation to exclude other causes of CPP; transabdominal and transvaginal duplex ultrasound (DUS) examinations of the pelvic venous system were performed using standardized protocols. Ultrasound probes with frequencies ranging from 3.5 to 5 MHz and 7 MHz were utilized for transabdominal assessment, whereas a 7.5-MHz transvaginal probe was used for detailed pelvic evaluation. The inferior vena cava, as well as the renal, gonadal, iliac, parametrial, and uterine veins, were systematically assessed regarding vessel diameter, patency, and the presence and duration of venous reflux. Reflux lasting longer than 0.5 seconds was considered pathologic. Additionally, the diameter and flow velocity of the iliac veins were measured. Comprehensive DUS examination of the lower limb and perineal veins was also performed to detect any abnormalities in the superficial or deep venous systems of the lower extremities; computed tomography venography or magnetic resonance venography were performed to exclude other causes of CPP; and pelvic venous reflux was confirmed by observing reverse flow during the Valsalva maneuver, venous dilation greater than 6 mm, and delayed contrast washout exceeding 5 seconds. Patients were asked to rate their average pelvic pain over the previous week using a Visual Analog Scale (VAS) ranging from 0 (no pain) to 10 (worst imaginable pain). Under local anesthesia, vascular access was achieved either via the femoral vein or the jugular vein using ultrasound guidance under sterile conditions, with the patient positioned supine. A curved stiff guidewire (0.035 inch, 260 cm) was introduced under C-arm fluoroscopic guidance. Over this wire, a Bernstein catheter (5F, 100 cm) was advanced to selectively cannulate the left renal vein. Venography of the left renal vein, the left common iliac vein, and the internal iliac (hypogastric) vein to rule out compression syndromes as a secondary cause of pelvic congestion. Once the left renal vein was assessed, the left ovarian vein was catheterized. A selective venogram of the left ovarian vein was performed to delineate its anatomy, identify all relevant collateral pathways, and evaluate for reflux and vein diameter while the patient performed a Valsalva maneuver. Incompetent ovarian veins were diagnosed according to established criteria, which included dilation of the ovarian vein with a diameter greater than 6 mm, documented reflux, contrast stasis within the pelvic plexuses for 5 seconds or more after injection, and abnormal venous connections between bilateral pelvic varices. In some cases, contrast descending to the perineum and genitalia was also observed, indicating extensive reflux pathways ( Fig 1 ). Fig 1 (A) Fluoroscopic venogram shows cannulation of ovarian vein. (B) Direct subtraction venogram shows ovarian vein. (C) Direct subtraction venogram shows ovarian vein and pelvic varicosities. (A) Fluoroscopic venogram shows cannulation of ovarian vein. (B) Direct subtraction venogram shows ovarian vein. (C) Direct subtraction venogram shows ovarian vein and pelvic varicosities.

From 40 cases of PCS due to ovarian vein reflux, the outcomes according to demographic data, causes of PCS, clinical picture of PCS, and vulvo-perineal or lower limb varicosities score are enumerated in Table I and Table II . Table I Demographic data and causes of pelvic congestion syndrome ( PCS ) Characteristics N = 40 Age, years 31.9 ± 5.99 Weight, kg 74.6 ± 11.57 Height, cm 162.9 ± 4.47 BMI, kg/m 2 28.2 ± 4.76 Parity 3 (1-3) Causes of PCS  Valveless ovarian vein 10 (25)  Pregnancy history 27 (67.5)  Hormonal factors 18 (45)  Genetics 9 (22.5) BMI , Body mass index. Data are presented as number (%), mean ± standard deviation, or median (interquartile range). Table II Signs and symptoms of pelvic congestion syndrome ( PCS ) and vulvo-perineal or lower-limb varicosities scores of the studied patients Signs and symptoms of PCS and scores Pain 40 (100) Vulvo-perineal or lower-limb varicosities 37 (92.5) Abdominopelvic tenderness 32 (80) Vaginal discharge 13 (32.5) Vulvar swelling 10 (25) Dyspareunia 22 (55) Dysmenorrhea 27 (67.5) Generalized lethargy 10 (25) Anxiety and depression 16 (40) Urinary frequency 13 (32.5) Data are presented as number (%). Demographic data and causes of pelvic congestion syndrome ( PCS ) BMI , Body mass index. Data are presented as number (%), mean ± standard deviation, or median (interquartile range). Signs and symptoms of pelvic congestion syndrome ( PCS ) and vulvo-perineal or lower-limb varicosities scores of the studied patients Data are presented as number (%). Patients with ovarian vein diameters >8 mm demonstrated significantly larger baseline measurements compared with those with diameters of 6 to 8 mm ( P < .001). Postprocedural analysis showed a reduction in ovarian vein diameter and pelvic varicosities in both groups ( Table III ). This reduction was less pronounced in the foam sclerotherapy group among patients with vein diameters >8 mm compared with the other subgroups ( P 8-mm group (n = 10) 6- to 8-mm group (n = 10) >8-mm group (n = 10) Preprocedure 7 ± 0.82 9.6 ± 0.52 7.4 ± 0.7 9.8 ± 0.42 <.001   P 1 <.001 .499 <.001   P 2 <.001 <.001   P 3 <.001 Postprocedure 2.3 ± 0.95 5.8 ± 1.48 2.1 ± 0.57 2.7 ± 1.06 <.001   P 1 <.001 .499 <.001   P 2 <.001 <.001   P 3 <.001 P1 , Comparison between pre- and postprocedure values; P2 , comparison between treatment groups; P3 , comparison between subgroups based on vein diameter. Data are presented as mean ± standard deviation. Boldface P values indicate statistical significance ( P < .05). Diagnosis by computed tomography of the studied subgroups P1 , Comparison between pre- and postprocedure values; P2 , comparison between treatment groups; P3 , comparison between subgroups based on vein diameter. Data are presented as mean ± standard deviation. Boldface P values indicate statistical significance ( P < .05). Recurrence (due to right ovarian vein) and vulvo-perineal or lower limb varicosities score were insignificantly different between both groups. Symptom improvement was not significantly different between both groups ( Table IV ). Table IV Outcomes and symptom improvement of the studied groups Group A (n = 20) Group B (n = 20) P value Outcomes  Recurrence 1 (5) 1 (5) – Symptom improvement  Complete relief 13 (65) 18 (90)  Partial improvement 7 (35) 2 (10) .058  No improvement 0 (0) 0 (0) Data are presented as number (%). Outcomes and symptom improvement of the studied groups Data are presented as number (%). Recurrence was insignificantly different among the four groups. Symptom improvement was significantly different between group B (6-8 mm and >8 mm) and group A (6-8 mm and >8 mm) ( P = .007) ( Table V ). Table V Outcomes and symptom improvement of the studied subgroups Group A (n = 20) Group B (n = 20) P value 6- to 8-mm group (n = 10) >8-mm group (n = 10) 6- to 8-mm group (n = 10) >8-mm group (n = 10) Recurrence 0 (0) 1 (5) 0 (0) 1 (5) .551 P 1 <.001 .977 .200 P 2 <.001 <.001 P 3 <.001 Symptom improvement  Complete relief 9 (90) 4 (40) 10 (100) 8 (80) .007  Partial improvement 1 (10) 6 (60) 0 (0) 2 (20)  No improvement 0 (0) 0 (0) 0 (0) 0 (0) P1 , Comparison between pre- and postprocedure values; P2 , comparison between treatment groups; P3 , comparison between subgroups based on vein diameter; VAS , Visual Analog Scale. Data are presented as number (%). Boldface P values indicate statistical significance ( P < .05). Improvement is defined as: complete improvement: above 90% reduction in VAS scale; partial improvement: from 50% to 90% reduction in VAS scale; and no improvement: below 50% reduction in VAS scale. Outcomes and symptom improvement of the studied subgroups P1 , Comparison between pre- and postprocedure values; P2 , comparison between treatment groups; P3 , comparison between subgroups based on vein diameter; VAS , Visual Analog Scale. Data are presented as number (%). Boldface P values indicate statistical significance ( P < .05). Improvement is defined as: complete improvement: above 90% reduction in VAS scale; partial improvement: from 50% to 90% reduction in VAS scale; and no improvement: below 50% reduction in VAS scale. VAS scores at baseline, 1 month, 3 months, 6 months, and 12 months were insignificantly different between both groups. Complications such as coil migration, hematoma at puncture, infection, hypersensitivity, and pulmonary embolism were insignificantly different between both groups ( Table VI ). Infection and pulmonary embolism did not occur in any patients in either group ( Table VII ). Table VI Visual Analog Scale ( VAS ) and complications of the studied groups VAS/complications Group B (n = 20) Group A (n = 20) P value VAS  Baseline 8 (7.75-9) 8 (7-9) .738  1 month 4 (3.75-5) 4 (3.75-5) .779  3 months 2 (1-4) 1 (1-2) .127  6 months 1.5 (0-4) 1 (0-2) .369  12 months 0 (0-4) 0 (0-0) .157 Complications  Coil migration 0 (0) 1 (5) 1  Hematoma at puncture 1 (5) 1 (5) –  Infection 0 (0) 0 (0) –  Hypersensitivity 1 (5) 0 (0) 1  Serious pulmonary embolism 0 (0) 0 (0) – Data are presented as median (interquartile range) or number (%). Table VII Complications of the studied subgroups Complications Group A (n = 20) Group B (n = 20) P value 6- to 8-mm group (n = 10) >8-mm group (n = 10) 6- to 8-mm group (n = 10) >8-mm group (n = 10) Coil migration 0 (0) 0 (0) 1 (10) 0 (0) .379 Hematoma at puncture 0 (0) 1 (10) 1 (10) 0 (0) .550 Infection 0 (0) 0 (0) 0 (0) 0 (0) – Hypersensitivity 0 (0) 1 (10) 0 (0) 0 (0) .379 Serious pulmonary embolism 0 (0) 0 (0) 0 (0) 0 (0) – Data are presented as number (%). Visual Analog Scale ( VAS ) and complications of the studied groups Data are presented as median (interquartile range) or number (%). Complications of the studied subgroups Data are presented as number (%).

CPP is described as cyclic or noncyclic lower abdominal and pelvic pain for more than or equal to 6 months. The pain might be from multiple factors with no evident cause, and its diagnosis and treatment may need the contribution of multiple specialties. 11 PCS is one of the most common causes of CPP and results from reflux or obstruction of the gonadal, gluteal, or periuterine veins, sometimes associated with perineal or vulvar varices. 12 , 13 In the current study, regarding causes of PCS, valveless ovarian vein was found in 10 patients (25%), pregnancy history was present in 27 patients (67.5%), hormonal factors were identified in 18 patients (45%), and genetic factors were involved in nine patients (22.5%). Aligned with our findings, Bałabuszek et al 14 stated that the high number of pregnancies, anomalies in pelvic venous anatomy, history of pelvic pain in family, and hormonal disorders like increased levels of estrogens were the risk factors for PCS. Along with our findings, Perry 15 noted that, regarding PCS, 13% to 15% of women lack valves in the left ovarian vein; the corresponding figure for the right vein was 6%. In the present study, regarding symptoms of PCS, pelvic pain was present in all patients, abdominopelvic tenderness was noted in 32 patients (80%), varicose ovarian vein in 21 patients (52.5%), vaginal discharge in 13 patients (32.5%), vulvar swelling in 10 patients (25%), and dyspareunia was reported in 22 patients (55%), whereas dysmenorrhea was observed in 27 patients (67.5%). Generalized lethargy in 10 patients (25%) and pelvic tenderness in 14 patients (35%) were also reported; anxiety and depression were present in 16 patients (40%) and urinary frequency in 13 patients (32.5%). Supporting our findings, Bałabuszek et al 14 mentioned that the prevalence of vulvar varices in patients with PCS was as high as 24% to 40%. Also, Whiteley 16 demonstrated that most patients with PCS develop CPP from reflux. Our findings revealed that ovarian vein diameter was significantly higher preprocedurally in patients with diameters >8 mm compared with those with 6 to 8 mm ( P < .001). Consistent with this, Barros et al 17 reported that pelvic vein diameters were larger in refluxing than in nonrefluxing veins. A significant postprocedural reduction in ovarian vein diameter was observed in both groups, especially group A (6-8 mm) and group B (6-8 mm and >8 mm). Also, Gava et al 18 demonstrated that foam sclerotherapy injection reduced the caliber of pelvic varicose veins, which leads to a significant reduction in complaints of pelvic pain among the patients after the procedure. In this study, recurrence (due to right ovarian vein) and vulvo-perineal or lower limb varicosities scores were insignificantly different between both groups. Recurrence was insignificantly different among the four groups. Vulvo-perineal or lower limb varicosities score was significantly improved in group A (6-8 mm) and group B (6-8 and >8 mm) than group A (>8 mm). These findings suggest comparable overall efficacy in preventing recurrence, whereas the sandwich technique may offer better varicosity resolution in larger veins. Consistent with this, Liang et al 19 concluded that treatment for varicose veins using a sandwich technique can be safe and efficient. Supporting our findings, Chen et al 20 reported that lower limb varicosities score did not differ significantly between both groups. According to this study, symptom improvement was insignificantly different between both groups. Symptom improvement was significantly different between group B (6-8 mm and >8 mm) and group A (6-8 mm and >8 mm). VAS at baseline, 1 month, 3 months, 6 months, and 12 months were insignificantly different between both groups. This suggested that, although both approaches offer comparable general pain relief, the sandwich technique may provide superior symptom reduction in patients with larger or more severely incompetent ovarian veins. Supporting our findings, Wong et al 21 found that 75% of patients reported symptom improvement following ovarian vein embolization. Furthermore, 13.8% of patients noted a complete resolution of their symptoms post ovarian vein embolization, and 37.5% of patients reported significant improvement. Consistent with our findings, Liang et al 19 reported that refluxing veins were treated using the “keep it simple” approach, deploying a minimal number of coils ‘sandwiching’ sclerosing foam. They reported that clinical success was 89% at 6 weeks and 84% at 1 to 5 years. VAS score reduction was achieved postprocedure. Our findings demonstrated that complications such as coil migration, hematoma at puncture, infection, hypersensitivity, and pulmonary embolism were insignificantly different between both groups. Coil migration, hematoma at puncture, and hypersensitivity were insignificantly different among the four groups. Infection and pulmonary embolism did not occur in any patients in either group A or group B. These findings indicate that both treatment techniques demonstrated comparable safety profiles. In line with our findings, Wong et al 21 concluded that ovarian vein embolization emerges as a safe and effective intervention for alleviating symptoms in patients with PCS. Also, Liang et al 19 concluded that there were no instances of coil dislodgement or other complications using the “keep it simple” approach, deploying a minimal number of coils ‘sandwiching’ sclerosing foam. This study has several limitations. It was conducted at a single center with a relatively small sample size, and no formal sample size calculation was performed, which may limit statistical power. Additionally, quality-of-life scores were not assessed, and the follow-up period was relatively short.

Both foam sclerotherapy and the sandwich coil embolization technique are safe and effective endovascular options for the management of ovarian vein reflux in patients with PCS. Overall outcomes, including pain relief, symptom improvement, recurrence rates, and complication profiles, were comparable between the two techniques. However, the sandwich technique demonstrated greater effectiveness in patients with larger ovarian vein diameters (>8 mm), whereas foam sclerotherapy provided satisfactory outcomes in smaller veins. These results support tailoring treatment according to ovarian vein diameter, with foam sclerotherapy serving as a suitable first-line option in selected patients, particularly in resource-limited settings.

After confirming the diagnosis of ovarian vein incompetence, a metallic coil (Interlock 35 detachable coils; Boston Scientific) was deployed distally in the ovarian vein segment. Following proximal coil placement, foam sclerotherapy was performed. The sclerosing agent used was polidocanol 3%, which was prepared as foam using the Tessari method. The foam was slowly injected under fluoroscopic guidance in a controlled manner to ensure complete filling of the refluxing segment without systemic embolization by an inflated 5F balloon (5 × 80 mm) to avoid any retrograde leak into renal vein. The maximum volume of polidocanol 3% foam injected did not exceed 10 mL per procedure. After foam injection, a second coil was deployed proximally in the ovarian vein to complete the sandwich occlusion. This method aimed to ensure both proximal and distal control, preventing foam or thrombus migration. A final venogram was performed to confirm the complete occlusion of the ovarian vein and absence of reflux ( Fig 2 ). Fig 2 (A) Fluoroscopic venogram shows distal coil in place. (B) Direct subtraction venogram shows sclerotherapy injection with distal coil in place. (C) Direct subtraction venogram shows total occluded ovarian vein with two coils in place (sandwich technique). (A) Fluoroscopic venogram shows distal coil in place. (B) Direct subtraction venogram shows sclerotherapy injection with distal coil in place. (C) Direct subtraction venogram shows total occluded ovarian vein with two coils in place (sandwich technique). After selective catheterization of the ovarian vein, as described previously, polidocanol 3% foam was prepared using the Tessari method and slowly injected under fluoroscopic monitoring by an inflated 5F balloon (5 × 80 mm) to avoid any retrograde leak into the renal vein. The injection continued until complete filling of the vein was achieved and no further reflux was visualized. After 5 minutes of injection, final venography was conducted postinjection to assess vein closure and confirm absence of residual reflux or unembolized segments, confirmed by the absence of contrast filling in the left ovarian vein and pelvic varicosities. Throughout both techniques, patients were closely monitored to detect any signs of nontarget embolization, hypersensitivity, or other procedural complications ( Fig 3 ). Fig 3 (A) Fluoroscopic venogram shows injection sclerotherapy of distal part of ovarian vein and pelvic varicosities by an inflated balloon. (B) Direct subtraction venogram shows occluded distal part of ovarian vein and partial occluded pelvic varicosities. (C) Fluoroscopic venogram shows injection sclerotherapy of proximal part of ovarian vein and remaining pelvic varicosities by an inflated balloon. (D) Direct subtraction venogram shows totally occluded ovarian vein and pelvic varicosities. (A) Fluoroscopic venogram shows injection sclerotherapy of distal part of ovarian vein and pelvic varicosities by an inflated balloon. (B) Direct subtraction venogram shows occluded distal part of ovarian vein and partial occluded pelvic varicosities. (C) Fluoroscopic venogram shows injection sclerotherapy of proximal part of ovarian vein and remaining pelvic varicosities by an inflated balloon. (D) Direct subtraction venogram shows totally occluded ovarian vein and pelvic varicosities. Every patient was thoroughly evaluated postprocedure to ensure the safety of the intervention and to monitor for potential complications. The puncture site was inspected regularly for hematoma formation or bleeding. Careful monitoring for signs of infection, including fever, localized redness, or swelling at the access site, was maintained throughout the hospital stay. Vessel perforation risk was evaluated intra- and postprocedure by reviewing final venographic images and monitoring patients for any new abdominal pain or hemodynamic instability. Pulmonary embolism was considered a rare but serious complication; therefore, all patients were monitored for respiratory symptoms. Hypersensitivity reactions to contrast media or sclerosants were documented and managed. Patients were routinely observed in the hospital for a minimum of 6 hours postprocedure. During this period, vital signs, oxygen saturation, pain scores, and puncture site conditions were closely monitored. Ambulation was encouraged as soon as possible to reduce thromboembolic risks. If patients remained stable with no signs of major complications, they were discharged on the same day with clear instructions. Prophylactic antibiotics were administered as per institutional protocol to minimize infection risk, typically a single preprocedural dose of a broad-spectrum antibiotic. Additionally, patients were prescribed analgesics and instructed on activity restrictions at home. Scheduled follow-up visits included clinical assessment and DUS to evaluate for late complications such as delayed coil migration, recurrent reflux, or formation of new varicosities. Patients were also instructed to report any symptoms such as pelvic pain, limb swelling, fever, or respiratory complaints immediately. We performed a comparison of each symptom’s intensity and signs (pelvic, genital, or lower limb) before and after the intervention. Pain was measured using VAS. Patients were categorized as either complete relief, partial improvement, or no improvement of symptoms and signs. We also performed an evaluation of vulvo-perineal or lower limb varicosities, an evaluation of recurrence of symptoms after improvement, and pelvic and transvaginal DUS (diameter and reflux). Patients were scheduled to follow-up at baseline, 1 month, 3 months, 6 months, and 1 year after the procedure. Statistical analysis was done by SPSS v29 (IBM Inc). Shapiro–Wilks test and histograms were used to evaluate the normality of the distribution of data. Quantitative parametric variables were presented as mean and standard deviation and compared between the two groups utilizing the unpaired Student t -test. Quantitative nonparametric data were presented as median and interquartile range and were analyzed by the Mann–Whitney test. Qualitative variables were presented as frequency (%) and were analyzed utilizing the χ 2 test or Fisher exact test when appropriate. A two-tailed P value ≤ .05 was considered statistically significant.

None.