Preliminary application of three-dimensional venography and fusion navigation technique in May-Thurner syndrome

Abstract

Objective

The purpose of this study was to report a technique for intraprocedural guidance of endovascular iliac vein stenting procedures using three-dimensional (3D) venography images as an overlay on live biplanar fluoroscopy.

Methods

Using 3D venography and a fusion navigation technique, percutaneous transluminal angioplasty and stent placement were performed to evaluate the feasibility of using 3D venography images and the fusion navigation technique to treat MTS compared with traditional digital subtraction angiography. The general epidemiologic data (ie, age, gender), clinical manifestations (ie, major symptoms, affected extremity, CEAP [clinical, etiology, anatomy, pathophysiology] classification, comorbidity, stenosis rate), intraoperative findings (ie, stent type, stent count, stent to inferior vena cava distance, procedure time, radiation dose, contrast agent dosage), and postoperative recovery were obtained and analyzed.

Results

A total of 30 consecutive patients with symptomatic MTS from our institution were enrolled in the present study. Of the 30 patients, 12 (group A) were treated using 3D venography images and fusion navigation and 18 (group B) were treated with two-dimensional venography images during endovascular management. Significant differences were observed between the two groups with respect to the procedure time (64.42 ± 4.35 minutes vs 76.61 ± 3.47 minutes; P = .04), radiation dose (2152 ± 124.7 mGy vs 2561 ± 105.6 mGy; P = .02), and contrast agent dosage (71.42 ± 4.87 mL vs 86.17 ± 4.14 mL; P = .03).

Conclusions

3D venography and its fusion navigation technique can improve prediction of the coverage area of the stent. Its use can also shorten the procedure time and reduce the contrast agent dose and radiation exposure, making it a valuable tool for both the diagnosis and the treatment of symptomatic MTS.

Article Highlights.

• •Type of Research: A single-center analysis of prospectively collected data
• •Key Findings: In the present study, three-dimensional venography images for fusion navigation was a feasible method and reduced the procedure time, radiation dose, and contrast agent dose in the treatment of May-Thurner syndrome.
• •Take Home Message: Three-dimensional venography and its auxiliary navigation technology enable the diagnosis and treatment of May-Thurner syndrome.

May-Thurner syndrome (MTS), also termed Cockett syndrome, is characterized by compression of the left common iliac vein by the right common iliac artery and against the fifth lumbar vertebra.^1,2 The clinical presentation of MTS is variable and diverse, with adult women the most affected.³ Treatment focuses on improving venous outflow from the affected limb. Stent placement is a safe and effective method for treating MTS.⁴ With the development of stenting came the use of venography, followed by intravascular ultrasound (IVUS), for the diagnosis and treatment of MTS. The current paradigm for confirmation of the diagnosis and treatment of MTS is via IVUS. However, IVUS has not been widely used in China.

Cone-beam computed tomography (CT) is a well-established imaging modality that has recently been integrated into C-arm angiography.^5,6 Cone-beam CT is based on the rotational movement around the patient of the C-arm equipped with a flat-panel detector. The acquired images are postprocessed to form a three-dimensional (3D) image for endovascular procedures and image-guided therapy. The common 3D images are derived from intracranial arterial 3D angiography, abdominal aortic aneurysm CT angiography, and interventional arterial 3D angiography. These 3D image fusion navigation techniques are used in the interventional treatment of malignant hepatic tumors, intracranial angiopathy, and so forth.^7,8 However, the venous approach to 3D reality is rarely applied. Therefore, we aim to use 3D venography images in the management of MTS. Because 3D venography provides CT-like images, it can be fused with live biplanar fluoroscopy to navigate the endovascular management flowchart for MTS; thus, endovascular management of all MTS cases is performed in an interventional suite. The type of endovascular treatment of MTS has not been well reported. In the present study, we evaluated the feasibility of 3D venography images and the fusion navigation technique for the diagnosis of MTS compared with traditional digital subtraction angiography (DSA).

Methods

Patients

The research ethics committee of our hospital approved the present prospective, observational, randomized cohort study. At our institution, we do not routinely measure the venous pressure because we consider such measurements to be of less value for patients in the supine position. The inclusion criteria were (1) lower extremity deep vein anterograde angiography via dorsal foot venipentesis showing iliac vein compression, which was preliminarily considered an iliac vein compression lesion; and (2) a CEAP (clinical, etiology, anatomy, pathophysiology) grade of C3 to C6. In addition, combined with a lesion and a normal vein, the diagnosis of MTS was determined by the presence of >50% area stenosis of the iliac vein with the establishment of venous collateral vessels. The exclusion criteria were as follows: (1) a history of serious trauma or major surgery within the previous 4 weeks; (2) pregnancy; (3) uncontrollable hypertension (systolic blood pressure >180 mm Hg or diastolic blood pressure >110 mm Hg); (4) a history of cerebral hemorrhage within the past 3 months; (5) life expectancy <1 year; (6) a contraindication to iodinated contrast media; (7) severe heart, liver, or kidney insufficiency; and (8) a history of deep vein thrombosis.

A total of 30 consecutive patients with symptomatic MTS (CEAP class >C3) from our institution were enrolled in the present study from October 2019 to October 2021. Of the 30 patients, 12 (group A) were treated with 3D venography images and fusion navigation, and 18 (group B) were treated with two-dimensional (2D) venography images during endovascular management. The general demographic data (ie, age, gender), clinical manifestations (ie, major symptoms, affected extremity, CEAP class,⁹ comorbidity, stenosis rate), intraoperative findings (ie, stent type, stent count, stent to inferior vena cava distance, procedure time, radiation dose, contrast agent dosage), and postoperative recovery were obtained and analyzed.

Image capture

In the present study, the GE IGS530 DSA machine (GE Healthcare, GE AW4.7 postprocessing station (GE Healthcare), and Mark 07 series high-pressure syringes (Medrad Inc) were used. The contrast agent used was iodixanol at a dose of 320 mg/mL (Hengrui Pharmaceuticals Co, Ltd).

Traditional DSA images were captured with the patient in the supine position. The Seldinger technique was used to achieve femoral vein access by introducing a 6F sheath (Terumo Corp) under local anesthesia and monitored anesthesia care. Under radiographic guidance, a 5F pigtail catheter (Terumo Corp) was positioned in the external iliac vein. The contrast agent, iodixanol (diluted to a concentration of 185 mg/150 mL), was injected using a high-pressure syringe with a pressure of 600 psi, an injection rate of 7 mL/s, and an injection volume of 15 mL. Anteroposterior and lateral projection venography images were obtained with the patients holding their breath.

3D venography images were captured. A 5F pigtail catheter was placed at the external iliac vein for venography to determine the lesion anatomy. The C-arm fluoroscopy mode was set to 3D-CT. The following parameters were set: contrast agent injection rate, 4 mL/s; total contrast volume, 8 mL; pressure, 300 psi; and delayed exposure, 0.5 second. 3D rotational venography was performed, during which the appropriate rotation degree (200°; 40°/s), ray delay time (1.5 seconds), contrast agent (4 mL/s; total volume, 24 mL; pressure, 300 psi) and other C-arm machine parameters were set. After venography, the data were transferred to a postprocessing workstation, and the 3D image was reconstructed using Innova 3D software (GE Healthcare). The images were postprocessed to clarify the stent coverage zones and avoid covering critical branch openings.

Navigation output

Output navigation can be processed using two methods: (1) the postprocessed 3D reality can be directly output to the navigation screen via the “export” option button; or (2) the lesion site can be observed from multiple angles and directions on the 3D images to select the clearest iliac vein stenosis angle, and the 3D reality can be sent to the C-arm using the “send angulation” option button. The operator was able to reach the desired angle for each frame quickly using the shortcut key. The device automatically generates 3D realities of various shapes in navigation, including dilution reality, virtual reality, and outline (Fig 1).

Fig 1.

Three-dimensional (3D) venography reality overlaid on fluoroscopy during an iliac vein stenting procedure. Virtual reality (A), dilute reality (B), and outline volume-rendered (C) views of the overlay.

Image matching and navigation

Iliac venography was performed again with guidance from the image output to the navigation screen. An appropriate 3D reality was selected to match the radiographic fluoroscopic image. Once the matching was completed, intelligent navigation was hosted by the device. The image automatically tracked in real time the position changes of the frame, C-arm, operating table, and flat plate detector to maintain accurate navigation. Using 3D reality and live biplanar fluoroscopy-fused navigation, a 0.35-in.guidewire (Glidewire; Terumo Corp) was directed through the stenosis of the iliac vein after percutaneous access to the ipsilateral common femoral vein was obtained. Percutaneous transluminal angioplasty using a Mustang balloon (diameter, 12-16 mm; length, 40-80 mm; Boston Scientific) and placement of Wallstents (Boston Scientific) were performed to treat underlying iliocaval compression. Self-expanding stents with a 20% larger than normal segment were placed in the iliocaval and iliofemoral veins, covering the entire diseased venous segment. The length of the stent was determined by the length of the lesion, with the proximal landing zone >20 mm to avoid stent shortening and migration. The distal stent extended ∼1.0 cm into the proximal vena cava when the lesion segment was located in the distal portion or ostium of the compressed common iliac vein. Fused navigation was primarily used during percutaneous transluminal angioplasty to optimize the placement of the balloon’s or stent's proximal and distal landing positions. Venography after stent placement revealed restoration of the normal iliac vein caliber (Fig 2). Balloon postdilation of the stented vessel was performed to improve stent fixation in the venous lumen and prevent displacement and migration. Balloon angioplasty was used to treat residual stenosis. After stenting, all patients received antithrombotic therapy for 6 months to prevent venous thromboembolism. The anticoagulant regimen was oral rivaroxaban (Bayer Consumer Health) at a dose of 15 mg/d and oral aspirin (Bayer Consumer Health) at a daily dose of 100 mg, both for 6 months.

Fig 2.

Application of three-dimensional (3D) venography image fusion navigation technique to treat May-Thurner syndrome (MTS). The guidewire entered the inferior vena cava smoothly through the true lumen (A), and the vessels in the lesion zone were successfully expanded under navigation guidance (B). C, The coverage zone of the proximal and distal ends of the stenosis were accurately located with navigation guidance. D, Postoperative 3D reconstruction showing stent morphology and posterior osteophyte after stent implantation.

Observed parameters and calculation methods

The critical parameters observed included the iliac vein stenosis rate, presence of collateral veins, radiation dosage, procedure time, and contrast agent dose. Using the traditional DSA lateral view, the minimum distance between the anterior and posterior walls of the most compressed portion of the iliac vein was recorded as distance a. The anteroposterior calibers of the iliac vein proximally and distally to the compressed area were recorded as distance b and distance c, respectively. The iliac vein stenosis rate was calculated as follows: [1 − 2a²/(b² + c²)] × 100%. On the 3D venography images, the proximal to distal ends of the stenosis lesion were connected to form a central line. We used the “two click AVA” shortcut key to straighten the vessels and measure the normal and minimum compressed cross-sectional area of the vessels. The area stenosis percentage of the potential venous compression point was calculated as follows: [1 − (compressed cross-sectional area/normal cross-sectional area)] × 100% (Fig 3). The radiation dosage and procedure time for each 2D venography images and 3D venography fusion navigation procedure were recorded from the fluoroscopy machine reports. Each contrast dose was recorded in detail by the imaging technician.

Fig 3.

A, Proximal to distal ends of the stenosis lesion (red arrow) were connected to form a central line. The “two click AVA” shortcut key was used to straighten the vessels (B,E) and measure the normal and compressed cross-sectional areas of the vessels (C,D). The area stenosis percentage of the potential venous compression point was calculated as follows: [1 − (compressed cross-sectional area/normal cross-sectional area)] × 100%; thus, the area stenosis percentage equaled [1 − 17.7(D)/146.1(C)] × 100% = 87.89%. An area stenosis percentage of >50% was defined as significant venous compression.

Statistical analysis

Statistical analysis was performed using SPSS statistics, version 26 (SPSS Inc). Continuous variables are presented as the median and range in the case of a nonparametric distribution and were compared using the Mann-Whitney U test. Continuous variables are presented as the mean ± standard deviation in the case of a parametric distribution and were compared using the independent t test. Categorical variables were compared using the χ² test and Fisher exact test and are reported as frequencies and percentages. A P value ≤ .05 was considered significant.

Results

A total of 30 patients with MTS who received endovascular management were included in the present study. Of the 30 patients, 16 were women (53.33%). The median age was 58.47 years (range, 43-78 years). The major symptoms associated with symptomatic MTS in groups A and B included pain (50% vs 55.56%; P = .77) and swelling (50% vs 44.44%). Of the 30 patients, 3 (10%) had hypertension, 3 (10%) had diabetes mellitus, and 3 (10%) had hypercholesterolemia. Moreover, the lesions of the 30 patients were C3 in 8 patients (26.67%), C4 in 9 patients (30%), C5 in 8 patients (26.67%), and C6 in 5 patients (16.67%). The stenosis rate was 81.31% and 81.38% in groups A and B, respectively. The clinical features and imaging findings are summarized in Table I.

Table I.

Baseline patient characteristics^a

Variable	Total (n = 30)	Group A (n = 12)	Group B (n = 18)	P value
Age, years	58.47 (43-78)	57 (46-71)	59.44 (43-78)	.5
Female gender	16 (53.33)	7 (58.33)	9 (50)	.72
Major symptoms
Pain	16 (53.33)	6 (50)	10 (55.56)	.77
Swelling	14 (46.67)	6 (50)	8 (44.44)
Left extremity affected	30 (100)	12 (100)	18 (100)
CEAP class
C3	8 (26.67)	3 (25)	5 (27.78)	.87
C4	9 (30)	4 (33.33)	5 (27.78)	.74
C5	8 (26.67)	3 (25)	5 (27.78)	.87
C6	5 (16.67)	2 (16.67)	3 (16.67)
Comorbidity
Hypertension	3 (10)	1 (8.33)	2 (11.11)	.8
Diabetes mellitus	6 (20)	3 (25)	3 (16.67)	.58
Hypercholesterolemia	3 (10)	2 (16.67)	1 (5.56)	.32
Stenosis rate, %	81.35	81.31	81.38

CEAP, Clinical, etiology, anatomy, pathophysiology.

Data presented as median (range) or number (%), unless noted otherwise.

^aGroup A, three-dimensional venography and fusion navigation; group B, two-dimensional venography images.

The stent type was the Wallstent in both groups. Significant differences were observed between the groups A and B, respectively, in the procedure time (64.42 ± 4.35 minutes vs 76.61 ± 3.47 minutes; P = .04), radiation dose (2152 ± 124.7 mGy vs 2561 ± 105.6 mGy; P = .02), and contrast agent dosage (71.42 ± 4.87 mL vs 86.17 ± 4.14 mL; P = .03). The patients in both groups had a low prevalence of complications, including back pain (23.33%), stent thrombosis (3.33%), and minor bleeding (16.67%). However, the differences between the two groups were not statistically significant. The treatment findings of the patients are summarized in Table II.

Table II.

Treatment of May-Thurner syndrome (MTS)^a

Variable	Total (n = 30)	Group A (n = 12)	Group B (n = 18)	P value
Stent type, Wallstent	30 (100)	12 (100)	18 (100)	NA
Stent count	33	15	18	NA
Stent to IVC distance, mm	10.6 ± 0.41	10.67 ± 0.69	10.56 ± 0.53	.9
Procedure time, minutes	71.73 ± 2.89	64.42 ± 4.35	76.61 ± 3.47	.04
Radiation dose, mGy	2397.6 ± 87.57	2152 ± 124.7	2561 ± 105.6	.02
Contrast agent dosage, mL	80.27 ± 3.38	71.42 ± 4.87	86.17 ± 4.14	.03
Complications				.86
Back pain	7 (23.33)	3(25)	4 (22.22)
Stent thrombosis	1 (3.33)	0 (0)	1 (5.56)
Stent migration	0 (0)	0 (0)	0 (0)
Major bleeding	0 (0)	0 (0)	0 (0)
Minor bleeding	5 (16.67)	2 (16.67)	3 (16.67)
Follow-up time, months	14 (3-23)	14 (3-23)	13.5 (3-21)	.99

IVC, Inferior vena cava; NA, not applicable.

Data presented as number (%), number, mean ± standard deviation, or median (range).

^aGroup A, three-dimensional venography and fusion navigation; group B, two-dimensional venography images.

Discussion

The diagnosis of MTS is a prerequisite for effective treatment. MTS is estimated to occur in 2% to 5% of patients with lower extremity venous insufficiency, reaching 18% to 49% in patients with deep vein thrombosis.¹⁰ Clinically, because of the diverse clinical symptoms and signs of MTS, its accurate diagnosis is not straightforward. The current gold standard for the diagnosis of MTS and guiding endovascular therapy is IVUS.¹¹ Many have advocated for the use of IVUS for vessel sizing and stent selection.¹² However, IVUS is only used in a very few Chinese hospitals, and most medical centers, including our institution, are not qualified to use this equipment. In the present study, 3D venography, particularly with the curved planar reformat and cross-section of the vessel, seemed sufficient for the diagnosis of MTS. Theoretically, this calculation method applies the 2D area ratio to assess the iliac vein stenosis rate, and the principle is the same as that using IVUS. However, from a diagnostic standpoint, the determination of the overall accuracy of 3D venography vs IUVS in the diagnosis of MTS requires prospective data.

Reconstruction of the iliac venous system and restoration of intravascular blood flow through recanalization, angioplasty, catheter-directed thrombolysis, or a stent can significantly relieve symptoms and lower the prevalence of MTS.¹³ Iliac vein stenting has been increasingly used in the treatment of MTS. However, an adequate assessment of the location of the stenosis and delineation of the optimal landing zone are critical factors for the success of endovascular interventions. The stent-covered zone must be identified to prevent undertreated proximal lesions and avoid affecting the contralateral iliac vein. Also, sufficient distal landing zones must be identified to prevent omission of distal lesions that could lead to in-stent restenosis or occlusion.^14,15 The 3D venography model can clearly show stenosis of the iliac vein and the entrance of the inferior vena cava during intervention. We found that the 3D venography model combined with live fluoroscopy is feasible and useful for understanding the iliac vein anatomy and stent position during the procedure. We performed the present study to describe the technique for intraprocedural 3D venography guidance, our preliminary experience, and the clinical potential of this technique for interventional treatment of MTS.

Owing to the slow blood flow and high resistance, 3D venography is more difficult to perform than 3D arterial angiography. Because the iliac vein blood flow velocity is affected by abdominal pressure, the patients in our study were instructed in the breath-holding technique before intervention. Iliac venography was performed in the lesion area with a pigtail catheter, and the images were collected while the patients held their breath. The benefit of the overlaid 3D reality was that it could be seen at all biplane angulations without repeat registration modification. Compared with 2D venography images, the 3D model provides superior visualization of vascular structures.¹⁶ The intervention was completed for 12 patients using the 3D model paired with live fluoroscopy at our institution. Stent implantation was convenient and smooth during the procedure. In addition, the procedure time, radiation dose, and contrast agent dose using 3D venography and the fusion navigation technique were significantly lower than those using 2D venography images.

Study limitations

The present study has several limitations. Because of the relatively small sample size, a risk exists of a type II statistical error. Also, IVUS scans could have been carefully scrutinized for intraluminal changes. However, we could not perform IVUS because it is not available at our institution. Therefore, no IVUS-related data were included in the present study. Finally, our follow-up is limited in scope. We hope to continue to accumulate data regarding the long-term follow-up of these patients.

Conclusions

The use of 3D venography and its fusion navigation technique can predict the coverage area of the stent. It can also shorten the procedure time and reduce the contrast agent dose and radiation exposure, making it a valuable tool in the diagnosis and treatment of symptomatic MTS.

Author Contributions

Conception and design: YX, YC, YG, GC, XH, YS, WW, XW

Analysis and interpretation: YX, YC, YG, GC, XH, YS, WW, XW

Data collection: YX, YC, YG, GC, XH, YS, WW, XW

Writing the article: YX, YC, YG, GC, XH, YS, WW, XW

Critical revision of the article: YX, YC, YG, GC, XH, YS, WW, XW

Final approval of the article: YX, YC, YG, GC, XH, YS, WW, XW

Statistical analysis: YX, YC, YG, GC, XH, YS, WW, XW

Obtained funding: YX

Overall responsibility: XW

YX and YC contributed equally to this article and share co-first authorship.