Long-term clinical effectiveness of endovenous interventions for pelvic venous disorders

Conception and design: NM, GL, RK, SL, SN, PP Analysis and interpretation: NM, PP Data collection: NM, RK, SN, PP Writing the article: PP Critical revision of the article: NM, GL, RK, SL, SN, PP Final approval of the article: NM, GL, RK, SL, SN, PP Statistical analysis: NM, PP Obtained funding: RK, SL, PP Overall responsibility: PP

This study was funded by a grant from the Lakhanpal Vein Foundation .

We retrospectively reviewed prospectively collected data from October 2016 to October 2022, from our Office of the National Coordinator for Health Information Technology certified electronic medical record (Nextgen Healthcare Information Systems) at the Center for Vascular Medicine (CVM). Institutional Review Board approval for the investigation was obtained (Advarra). Informed consent was not required as per the Institutional Review Board. Women with a PeVD secondary to OVR alone, IVS alone, and combined IVS and OVR were included in the study. It is our protocol at CVM that all patients receive a thorough evaluation by a gynecologist for nonvenous-related causes of chronic pelvic pain before considering interventions for symptomatic PVI. Once other pelvic pain causes were ruled out, a CVM venous specialist evaluates patients for the presence of symptomatic PVI. A venous specialist is defined as a physician who dedicates their practice to the care of patients with venous disease. At CVM, this includes board-certified physicians in vascular surgery, cardiology, gynecology or cardiothoracic surgery. We also use Advanced Practice Nurses and Physician Assistants to optimize the intake process. All patients with a negative gynecologic assessment for chronic pelvic pain, a history of pelvic pain, dyspareunia, postcoital pain, urinary frequency, flank pain, abdominal bloating, and/or clinical evidence of lower-extremity pelvic escape veins, were evaluated with transabdominal ultrasound (TAU) study. Patients with clinical symptoms and positive TAU findings were offered, diagnostic venography, intravascular ultrasound (IVUS) study and possible iliac vein ST. Our TAU protocol has been previously published. 33 , 34 Duplex ultrasound study was performed preintervention, 1 week postintervention, 3, 6, 12, 18, and 24 months, and yearly thereafter. Medical/surgical histories, Clinical-Etiology-Anatomy-Pathophysiology clinical class and preintervention, 1, 3, 6, 12, 18, 24, and 36 months visual analog scale (VAS) pain scores, stent type, diameter, length, and vein territory covered, were all evaluated. All treated women underwent a diagnostic venogram which consisted of an anteroposterior view of the bilateral iliac veins, inferior vena cava, left renal vein, left ovarian veins, and any pelvic veins. Balloon-occlusion venography of the internal iliac veins and interrogation of the right ovarian vein were not performed. For patients requiring OVE, chemical embolization with 3% Sodium Tetradecyl Sulfate foam (Mylan Institutional LLC) or Interlock coils (Boston Scientific) were used as previously described. 8 All patients had pre- and post-ST IVUS study for identification of an area reducing lesion of the iliac veins and confirmation of good wall apposition and appropriate stent expansion. If a residual stenosis required postdeployment venoplasty, a completion IVUS study was performed to ensure proper expansion of the residual stenosis. A pelvic reservoir was defined as the presence of cross pelvic collaterals, the presence of at least three unilateral, and/or bilateral pelvic segments that communicated with either the internal iliac vein or ovarian veins. All venograms were reviewed by one of the senior authors (G.L.) for accuracy. Antiplatelet therapy poststent implantation is not used. Anticoagulation with a factor Xa inhibitor for 90 days poststent implantation is used at the discretion of the treating physician. Indications for extending anticoagulation past 90 days is the presence of thrombus layering as identified by duplex scanning, a residual stenosis or the presence of intraluminal post-thrombotic scar tissue. Patients were separated into groups on the basis of symptom presentation: leg pain only, pelvic pain only, and combined leg and pelvic pain. Time-to-maximal pain reduction and the effectiveness and durability of treatment strategies using VAS pain scores were subsequently analyzed. Treatment strategies analyzed were as follows: for patients with pelvic pain alone, OVE alone, and OVE with iliac vein ST; for leg pain alone, iliac vein ST was assessed; for combined disease, OVE with ST and iliac vein ST alone were assessed. Discriminative data were calculated using χ 2 analysis, and continuous categorical values were analyzed using analysis of variance. The statistical software program GraphPad Prism, version 10.0 (Boston) was used to perform all calculations.

From October 2016 to October 2022, 521 women with a PeVD were evaluated and treated. Patients were categorized according to symptom presentation in the following manner: pelvic pain alone (n = 33), leg pain alone (n = 208), and pelvic and leg pain (mixed; n = 280). Patients were treated with either OVE alone, OVE and ST, or ST alone. The average age of the cohort was 52 ± 13.4. Table outlines the demographics of the entire cohort. Patients with pelvic pain were younger than those in all other groups, and patients with leg pain were older than those with pelvic and mixed pain ( P ≤ .001). There were no differences in symptom presentation based on race. Patients with pelvic and leg symptoms had a higher prevalence of endometriosis and ovarian cysts ( P ≤ .05). Fig 1, A and B demonstrate the Clinical-Etiology-Anatomy-Pathophysiology classes of the right and left limbs for each presentation category. The majority of patients presented with C3 disease or less. Fig 2, A and B demonstrate the types of stents, diameters, lengths, and vein segments treated. A total of 313 stents were implanted: 180 Venovo (Becton Dickinson), 6 Wallstents (Boston Scientific), and 127 Abre (Medtronic). The most common stent diameters and lengths were 14 and 16 mm and 120 and 150 mm, respectively. The inferior vena cava, left common, and external iliac veins were the most common vein segments treated followed by the right common and external iliac veins. Table Demographics of the entire cohort Variables Total (N = 521) Pelvic pain (N = 33) Leg pain (N = 208) Mixed pelvic plus leg (N = 280) P value Age 52 ± 13.4 39 ± 13.6 59 ± 12.1 49 ± 12.1 .001 Sex  Female 521 33 208 279 .65  Race  Asian 6 0 1 5 1.00  Black/African 121 6 55 57 .23  Hispanic/Latino 63 4 15 43 .30  Indian 2 0 1 1 1.00  White 320 19 128 155 .39  Other race 2 0 0 2 1.00  Multiracial 2 0 0 2 1.00  Unknown 19 3 6 8 .11 Medical history  Fibroids 29 2 11 15 .98  Endometriosis 30 3 3 22 .01  Ovarian cysts 13 1 1 11 .05  PCOS 9 1 1 6 .26  Hysterectomy 6 0 1 4 .48 PCOS , Polycystic ovary syndrome. Fig 1 A and B, Clinical-Etiology-Anatomy-Pathophysiology ( CEAP ) clinical classification of the entire cohort: right limbs (A) and left limbs (B) . Fig 2 A, Types of stents, diameters, and lengths used to treat patients with a pelvic venous disorder (PeVD). B, Iliac vein segment territories covered by implanted stents. IVC , Inferior vena cava; LCFV , left common femoral vein; LCIV , left common iliac vein; LEIV , left external iliac vein; RCFV , right common femoral vein; RCIV , right common iliac vein; REIV , right external iliac vein. Demographics of the entire cohort PCOS , Polycystic ovary syndrome. A and B, Clinical-Etiology-Anatomy-Pathophysiology ( CEAP ) clinical classification of the entire cohort: right limbs (A) and left limbs (B) . A, Types of stents, diameters, and lengths used to treat patients with a pelvic venous disorder (PeVD). B, Iliac vein segment territories covered by implanted stents. IVC , Inferior vena cava; LCFV , left common femoral vein; LCIV , left common iliac vein; LEIV , left external iliac vein; RCFV , right common femoral vein; RCIV , right common iliac vein; REIV , right external iliac vein. The average preintervention VAS scores were as follows: among patients with pelvic pain treated with OVE plus ST and OVE alone, the VAS scores were 3.07 ± 3.71 and 5.27 ± 3.2, respectively. Among patients with leg pain treated with ST alone, VAS scores were 6.18 ± 3.07. Among patients with mixed pain treated with ST plus OVE and ST alone, the VAS scores were 7.47 ± 2.28 and 6.67 ± 2.56, respectively. Significant improvements in pain scores were observed in all treatment groups ( Fig 3, A-C ). Regardless of the symptom presentation group, patients who received OVE plus ST demonstrated significant pain reduction at 1 month with no further decrement in pain that was sustained up to 36 months ( P ≤ .001). Patients treated with either ST or OVE alone demonstrated maximal pain reduction at 3 months with no further pain decrement that was similarly sustained through 36 months ( P ≤ .001). Fig 3 A, Visual analog scale ( VAS ) pain scores for patients with pelvic pain treated with ovarian vein embolization ( OVE ) or OVE and stenting (ST) demonstrating sustained pain relief over 36 months with maximal pain reduction at 3 months for OVE alone and 1 month for OVE and ST. B, VAS pain scores for leg pain patients treated with ST demonstrating sustained pain relief over 36 months with maximal pain reduction at 3 months. C, VAS pain scores for patients with pelvic and leg pain demonstrating sustained pain relief over 36 months with maximal pain relief at 1 month for patients treated with OVE and ST and 3 months for patients treated with ST alone. A, Visual analog scale ( VAS ) pain scores for patients with pelvic pain treated with ovarian vein embolization ( OVE ) or OVE and stenting (ST) demonstrating sustained pain relief over 36 months with maximal pain reduction at 3 months for OVE alone and 1 month for OVE and ST. B, VAS pain scores for leg pain patients treated with ST demonstrating sustained pain relief over 36 months with maximal pain reduction at 3 months. C, VAS pain scores for patients with pelvic and leg pain demonstrating sustained pain relief over 36 months with maximal pain relief at 1 month for patients treated with OVE and ST and 3 months for patients treated with ST alone.

PeVDs are increasingly recognized as a major etiology for pelvic and/or leg pain. Broadly, pelvic vein anomalies present with either obstructive and/or reflux pathologies in the iliac, ovarian, hypogastric, or renal veins. This observation is the foundation of the Symptoms-Varices-Pathophysiology classification system. 35 We have previously reported that in patients with OVR and/or iliac vein outflow obstruction treated with OVE followed by ST at 1 month that little to no VAS pain reduction was observed with OVE. We have also reported that in patients with a PeVD secondary to reflux and obstruction, ST alone resulted in complete pain resolution in 78% of patients. 5 , 8 , 31 The criticism of our initial results with staged OVE and ST is that a longer interval between OVE and ST may have demonstrated a clinical benefit with OVE. Despite these concerns, our group has abandoned OVE in patients with combined reflux and obstruction and solely focus on relieving outflow disease. However, the unknown long-term consequences associated with iliac ST in younger patients has prevented widespread adoption of this treatment modality. Given the extensive degree of venous pathologies that cause pelvic pain, clinical decision making has focused on the extent and type of treatment needed to ameliorate the patient's pain issues. For patients whose pelvic symptoms are secondary to isolated ovarian/hypogastric vein reflux or isolated IVS with bilateral disease, it is currently unknown if limited ovarian/hypogastric embolization or limited ST interventions are sufficient to provide long-term and durable results. Numerous investigations have reported improvement with unilateral left ovarian vein disease; however, patients typically report some continuous pelvic pain. 10 , 17 , 20 , 22 , 23 , 24 , 25 , 27 , 28 , 36 Degregorio et al 7 reported that embolization of bilateral ovarian and hypogastric veins in patients with pelvic vein reflux alone demonstrated superior and sustained 6 year results. In patients with pelvic/leg symptoms who present with bilateral iliac vein stenoses, our group only treats the lesion with the highest area reducing lesion for pelvic pain alone or the lesion associated with ipsilateral leg symptoms. We subsequently observe the patient for symptom recurrence and noted that only 22% presented with a symptom recurrence from the untreated contralateral lesion over a 6-year period. It therefore appears that for reflux disease alone, aggressive elimination of all sources of reflux is associated with superior results, whereas limited ST with observation may decrease the number of stents placed and subsequent reinterventions in the future. In patients with iliac vein obstruction and ovarian/hypogastric vein reflux, the current literature suggests that obstructive lesions are the primary pain generators and that OVR in these patients is usually asymptomatic. 1 , 5 , 31 , 32 In patients with mixed disease treated with ST alone, several investigators have indicated that between 73% and 78% of patients reported complete symptom resolution. 5 , 8 , 31 Gavrilov et al 32 reported on 55 patients treated with mixed disease. Six patients were treated with OVE as the initial form of therapy, and all six demonstrated no improvement in symptoms. Symptom resolution was attained only after iliac vein ST. In the remaining 49 patients, all were treated with iliac vein ST as the initial therapy with 39 patients reporting symptom improvement (69.4%). The remaining 15 patients (30.6%) were treated 6 months after ST with OVE and reported significant symptom improvement. It must be noted that Gavrilov et al 32 do not utilize IVUS study to assess the degree of area reduction and solely rely on venography. The Vidio trialists demonstrated the unreliability of venography to accurately assess the degree of obstruction. 37 Therefore these 15 patients may have actually been primary reflux patients and emphasizes the necessity of accurate diagnostic imaging before intervention. These data clearly support the supposition that in patients with combined reflux and obstruction that obstructive lesions are the primary venous origin pain generators. Despite the evidence suggesting superior results with iliac vein ST in patients with mixed disease, most clinicians want to avoid placing a permanent, nonretrievable implant in a 20 to 40-year-old patient. No stent stays patent indefinitely. Placing a stent in a young patient requires lifelong clinical follow-up and potential reinterventions to avoid symptom recurrence or worsening outflow disease. Our group has previously reported a 14% reintervention rate up to 36 months in patients with mixed disease treated with ST alone, emphasizing the long-term risks of iliac vein ST. 1 The current investigation validates our group's initial observations in terms of VAS score reductions. The patients treated with OVE and ST had the greatest reduction in VAS scores and achieved maximal pain reduction the quickest at 1 month. Patients receiving OVE or ST alone took 3 months to achieve the same pain reduction. These observations clearly suggest that eliminating all sources of reflux and obstruction provide quicker results even though the same degree of pain reduction was observed with either approach. These data also suggest that there may be a subset of patients who benefit from treating a pelvic reservoir while simultaneously relieving any outflow disease. Some clinicians who advocate for an OVE approach first in mixed patients recommend waiting 6 months before considering reintervention or iliac vein ST. The current data set clearly indicates that waiting more than 3 months will not result in further pain reduction and only increases patient discomfort time. The question that needs further elucidation is whether or not OVE in these patients should be considered a first-line therapy before considering iliac vein ST. Gavrilov et al 32 tried this approach and abandoned it after treating six patients with OVE first. Symptom improvement was only observed with subsequent ST. The lack of significant pain reduction with OVE in these patients may be due to limited embolization and leaving disease behind. It is currently unknown whether an aggressive four-vessel embolization regimen like that used by De Gregorio et al 7 would result in pain reduction adequate enough for patient quality of life improvement as De Gregorio’s investigation excluded patients with iliac outflow disease. Complete pain resolution may not be necessary in these patients. If a VAS reduction to a level of three is satisfactory and obviates the need for an iliac vein stent, then serious consideration to an aggressive OVE first regimen needs investigation. The retrospective analysis of prospectively collected data, the low number of OVE alone patients, the use of unilateral left OVE and the reliance of VAS as the main outcome parameter are the major limitations of this study. After our 2018 publication we pursued a ST first approach and therefore limited the number of OVEs performed. It is clear from Degrogorio’s investigation that aggressive four-vessel embolization in patients with OVR without an outflow obstruction demonstrates superior and sustained pain reduction. The current investigation also clearly indicates that in patients with reflux and obstruction, a combined OVE and ST strategy may yield better results than OVE or ST alone. Therefore, our OVE results may have demonstrated improvement in pain reduction sooner than 3 months if a four-vessel embolization strategy was used. Finally, these data must be interpreted knowing the limitations of using VAS scores as the primary assessment parameter. The lack of a PeVD quality of life questionnaire and the variability in pain experienced by patients with PeVD on a day-to-day basis invariably exposes interpretations to bias.

Endovascular therapies for PeVD disorders are currently the standard of care. These data indicate that regardless of the treatment strategy, patients with PeVD demonstrate sustained pain reduction through 36 months. Patients treated with OVE and ST reported maximal pain reduction at 1 month, whereas ST or OVE alone took 3 months. No further decrement in pain was observed after 3 months. In patients wishing to avoid stent placement, a staged approach should be considered. If a staged approach with OVE followed by ST is used in patients with ovarian/hypogastric vein reflux and iliac vein obstruction, ST should be considered at 3 months as no further decrement in pain was observed after that time period.

The authors have no competing interests.