Abstract
Keywords
Introduction
Superficial chronic venous disease (CVD) has a high prevalence within the community1,2 and leads to significant patient morbidity, reduced quality of life, time away from work, and increased healthcare expenditure.3,4 It commonly results from valvular reflux in the saphenous vein and ambulatory venous hypertension, which may lead to a spectrum of venous disorders ranging from varicose veins to ulceration of the skin.5–8 While venous ligation and saphenous vein stripping were historically favored for the treatment of CVD, most venous specialists have now moved their practice toward venous ablation with laser, radiofrequency, or cyanoacrylate glue embolization to avoid large surgical incisions, reduce postoperative pain, and facilitate a faster recovery.9–12
These ablation procedures have traditionally relied on B-mode ultrasound imaging to guide the insertion and placement of the ablation catheter. 13 While ultrasound provides high-quality, real-time imaging of the vein and surrounding structures, its utility may be limited in anatomically complex cases. Challenges such as venous stenoses, duplicate saphenous systems, tortuosity, segmental occlusions, aneurysms, or a vein that lies deep within the leg may be incompletely imaged.
Digital subtraction venography (DSV) is the imaging modality of choice for a number of venous conditions, including pelvic and gonadal vein insufficiency,14,15 venous malformations, 16 and May-Thurner syndrome. 17 It gives clear and precise images that may have utility in the treatment of saphenous vein insufficiency to identify and negotiate challenging anatomical variations, facilitate selective cannulation when multiple saphenous trunks are encountered, and provide precise positioning of thermal ablation probes when ultrasound is challenged.
The aim of this study was to evaluate the utility of DSV in the identification of venous anomalies that may alter definitive patient management and facilitate successful treatment in cases where ultrasound alone may be insufficient.
Methods
Study Design
Data were gathered from 3 Australian medical centers to analyze the utility and efficacy of venography used adjunctively during thermal and nonthermal ablative procedures for saphenous vein insufficiency. Patients underwent baseline venous assessment with duplex ultrasound in a dedicated vascular diagnostic laboratory by experienced vascular sonographers. Sonographic reflux >0.5 seconds and CEAP (clinical, etiological, anatomical, and pathological) classifications 2–6 were indications for an ablative procedure with possible venography in the great, small, intersaphenous, or perforator vein. Pregnant patients, those with contrast allergy or renal impairment precluding the use of iodinated contrast were not eligible, as were those who underwent truncal vein ligation and stripping, simple phlebectomy, pelvic vein embolization, or interventional treatment for May-Thurner syndrome. The human research and ethics committees of the participating centers approved the retrospective study; all patients provided written informed consent for the procedure. Initial procedural and demographic data were collected prospectively with an electronic medical records program at each participating center; relevant data and reports were transferred to a database (Excel 2007; Microsoft, Redmond, WA, USA) for review and analysis.
Procedural images were independently evaluated by 2 vascular specialists experienced in the interpretation of digital subtraction angiography and intraprocedural venography to collect procedural data and target vessel characteristics, such as vein treated, presence of venous anomalies, and the use of fluoroscopy-guided maneuvers to effect complete venous ablation. Venograms were compared to preoperative duplex scans when available to determine the sensitivity, specificity, and positive and negative predictive values of ultrasound in the detection of abnormalities seen on venography. Discrepancies were adjudicated by consensus where possible or by a third specialist.
Patient Population
Between October 2010 and May 2016, 200 consecutive patients (mean age 60.9 years, range 33–86; 128 women) with CVD in 268 limbs underwent ablation guided by ultrasound and adjunctive venography and fluoroscopy of 305 saphenous, intersaphenous, and/or perforator veins. The patients’ baseline characteristics, CEAP classifications, and target veins are shown in Table 1.
Demographics and Disease Severity Scores in the 200 Study Patients. a
Abbreviations: AAGS, anterior accessory great saphenous; CEAP; clinical, etiological, anatomical, and pathological elements.
Continuous data are presented as the means ± standard deviation; categorical data are given as the counts (percentage).
Venous Procedures
All procedures were performed in an operating theatre equipped for both duplex ultrasound and DSV using a fixed, ceiling-mounted C-arm system (Allura Xper FD20; Philips Healthcare, Best, the Netherlands). The type and extent of each treatment were determined at the discretion of the attending vascular specialist.
Venous ablation was performed under either local or general anesthesia at the discretion of the operator and attending anesthetist. A micropuncture needle and sheath (Merit MAK Mini Access Kit; Merit Medical, South Jordan, UT, USA) were employed for ultrasound-guided venous access at a location distal to the segment of incompetent vein. After exchanging for a 7-F sheath (Cordis Corporation, Bridgewater, NJ, USA), DSV was performed to further define vein anatomy and the characteristics of the target vein. When anatomical complexities that prevented device passage along the target vein were encountered, the operator would first attempt to facilitate delivery with external manual compression, knee bending, or altered leg position. If those maneuvers failed, a low-profile guidewire and angled catheter were used to selectively cannulate the target vein, using digital subtraction roadmaps and/or fluoroscopy to guide manipulation of the wire to the deep vein junction.
If radiofrequency ablation was planned, a Venefit probe (Medtronic/Covidien, Minneapolis, MN, USA) was inserted and positioned at the superficial-deep vein junction under fluoroscopic guidance. Ultrasound-guided tumescent analgesia (500 mL normal saline, 30 mL lidocaine 2%, 5 mL sodium bicarbonate solution, and 0.5 mL of adrenaline 1:1000) was infiltrated around the saphenous vein prior to commencing radiofrequency energy. Two 20-second cycles at 120°C were initiated 1 to 2 cm distal to the junction followed by single cycles down to the vein access point in 6-cm segments.
If the operator opted for cyanoacrylate glue embolization, a Venaseal sheath (Sapheon; Medtronic/Covidien) was delivered 5 cm distal to the deep vein junction. Two 0.1-mL aliquots of cyanoacrylate glue were delivered 1 cm apart; the catheter was withdrawn 3 cm and then the vein compressed for 3 minutes to ensure complete polymerization. Following that, 0.1-mL aliquots were delivered and compressed for 30 seconds every 3 cm until the distal vein was completely treated.
After complete saphenous ablation within the thigh, phlebectomies were undertaken through microincisions, or sclerotherapy was performed where indicated. Hemostasis was achieved through manual pressure and the placement of a whole leg Eloflex compression bandage (8×125 cm; Smith & Nephew, Sydney, Australia) after steri-strips were applied to the microincisions. Patients were evaluated clinically prior to discharge and scheduled for routine outpatient follow-up at 1 month. Adverse events and complications within 30 days of the operation were recorded.
Definitions
Venous stenosis was defined as a >70% diameter reduction of the target vein compared with the adjacent normal reference segment. Aneurysm was a solitary area of venous dilatation >150% of the adjacent reference vein and not associated with an arteriovenous malformation, pseudoaneurysm, or varicose vein. 18 A large incompetent perforator was >4 mm in diameter and demonstrated flow in the direction of the deep to superficial saphenous vein. A duplicate saphenous vein was one that split into 2 or more interconnected trunks within the saphenous sheath. Other anomalies included communication with an anterior accessory great saphenous vein (AAGSV), segmental saphenous vein occlusion, and the presence of intraluminal venous thrombus.
Technical success was defined as the ability to deliver the ablation catheter to the target vein from the access point and deploy radiofrequency energy or cyanoacrylate glue the entire length of the incompetent segment, completely ablating the insufficient saphenous trunk.
Statistical Analysis
The Pearson chi-square or Kruskal-Wallis tests were used for categorical data analysis. Continuous data were summarized as mean ± standard deviation or median (range) and were compared using a parametric t test or 1-way analysis of variance. All tests were 2-sided, and p<0.05 was considered significant. Statistical analysis was performed using IBM SPSS software (version 22; IBM Corporation, Somers, NY, USA).
Results
Procedural Outcomes
The majority of patients (174, 87%) underwent radiofrequency ablation, while the remaining 26 (13%) had cyanoacrylate glue embolization. Technical success was 100%, and there were no contrast-related complications during the procedure, at discharge, or at the 30-day follow-up.
There were 542 venograms completed (2.0/limb) over a mean acquisition duration of 4.9±9.1 minutes (median 1 minute; range 1–48). Venous anomalies and abnormalities were present in 132 (66%) of the 200 patients (Table 2; Figure 1). Fluoroscopically-guided adjunctive maneuvers (Table 3, Figures 2 and 3) were required in 88 (44%); most common (27%) were the use of a low-profile guidewire and catheter to maneuver through tortuosity or stenosis or the use of additional radiofrequency energy or cyanoacrylate glue in particularly dilated venous segments (15%).
Target Vein Anomalies and Abnormalities Detected by Venography in 200 Patients. a
Abbreviation: AAGSV, anterior accessory great saphenous vein.
Data are given as the counts (percentage).
Commonly encountered venous anomalies and abnormalities. (A) An incompetent great saphenous vein (GSV) with dilated aneurysm (arrowhead) and a stenosed segment (arrow). (B) A duplicate GSV; arrowheads denote the incompetent branch and direction of reflux. (C) GSV with aneurysm (arrowhead) duplicated in the thigh. (D) An incompetent perforating vein in the thigh, which renders the distal GSV incompetent. The proximal GSV was competent on duplex ultrasound and nondilated on venography. (E) Partially occlusive thrombus within the GSV and tributaries, often observed after past sclerotherapy.
Adjunctive Maneuvers Made Possible in 200 Cases With Assistance of Fluoroscopy and/or Venography. a
Abbreviations: CGE, cyanoacrylate glue embolization; RFA, radiofrequency ablation.
Data are given as the counts (percentage).
(A) A great saphenous vein with segmental occlusion in the distal thigh and (B) tortuous tributary connecting the proximal and distal segments (arrow denotes the occluded great saphenous vein segment). (C) An angled catheter (arrow) and low-profile guidewire are used to negotiate the tortuous segment and (D) deliver the radiofrequency catheter to the saphenofemoral junction.
(A) A duplicate great saphenous vein with incompetent branch indicated by the arrow. (B) With an access sheath from below, the sharp takeoff (arrow) to the incompetent branch (asterisk) made it impossible to simply advance the radiofrequency probe. (C) An angled catheter (arrow) and low-profile guidewire are used to selectively cannulate the target vein (arrow). (D) Advancing the radiofrequency probe (arrow) over the guidewire under fluoroscopic guidance.
Operators deemed the use of adjunctive fluoroscopy/venography to be necessary for successful completion of the procedure in 34 (17%) cases. Recorded factors that prompted this judgment included severe tortuosity (13, 38%), stenosis preventing catheter passage (9, 27%), an occluded saphenous vein that required wire traversal to reach an incompetent proximal component (5, 15%), and the need for selective guidewire cannulation of a duplicate trunk (14, 41%).
A simultaneous phlebectomy procedure was performed in 191 (95.5%) of the 200 patients. The remainder underwent sclerotherapy (2.5%) or no localized treatment for varicose veins (2%). Surgical complications occurred in 11 (5.5%) patients at 30 days. These included saphenous neuropraxia/neuralgia (4%) and thrombophlebitis (1.5%). There was general symptomatic improvement or complete resolution of preexisting symptoms reported in 96.5% at the 30-day follow-up.
Comparison of Preoperative Duplex Ultrasound and Intraprocedural Venography
The majority of patients (193, 97%) who underwent intraprocedural venography had a preoperative venous duplex scan available for comparison. Evaluation of the reports from those studies identified 21 (11%) patients with abnormalities detected on ultrasound (23 anomalies) compared with 123 (64%) on venography (193 anomalies). This gave ultrasound a 17.1% sensitivity, 100% specificity and positive predictive value, and 40.7% negative predictive value. The preoperative duplex report was most likely to identify organized thrombus (2/6, 33%), followed by segmental occlusion (2/7, 29%), a significant communication with the AAGSV (3/11, 27%), a duplicate saphenous system (12/61, 20%), a large incompetent perforator vein (2/19, 11%), target vein stenosis (1/52, 2%), and venous aneurysm (0/41, 0%).
Discussion
This study determined the utility of DSV in identifying saphenous vein anomalies during the treatment of CVD patients so as to achieve complete vein ablation. In two-thirds of the patients, the abnormal saphenous vein findings included a number of anomalies with the potential to impede technical success and affect management decisions. These observations led the operator to utilize fluoroscopically-guided techniques to achieve complete treatment success of the correct target vein in 44% of cases. Moreover, ablation of the entire incompetent venous segment was deemed impossible with ultrasound alone in 17% of the cohort; success was achieved only with the aid of adjunctive x-ray guidance, clearly illustrating the practical value of x-ray imaging in ablating the length of the target vein.
Duplex ultrasound is the gold standard for the preoperative imaging of patients with suspected CVD.19,20 It has also been used as the adjunctive imaging modality of choice to visualize the treatment of saphenous veins during thermal and nonthermal ablation. 13 It is useful in facilitating percutaneous vein access, guiding the introduction of tumescent anesthesia, positioning the thermal catheter, and focusing extrinsic compression of the vein when required.
This study evaluated the incidence of anomalies reported in the preoperative duplex studies and compared them to the intraoperative venographic images to determine if the preoperative reports could be used to accurately identify the likelihood of encountering anatomy that would pose a technical challenge. Unexpectedly, preoperative ultrasound provided a low 17.1% sensitivity and 40.7% negative predictive value in the detection of those anomalies, contrary to expert opinion that high quality images can be attained with current generation ultrasound machines.
These findings may be explained by considering how a vascular sonographer performs a venous duplex study and reports the observed signals. The goal of these investigations is to provide a diagnosis and map of the underlying physiological dysfunction that correlates with the venous symptoms; the sonographer determines vein patency, valvular competence, and simple anatomical variants. The report produced is therefore likely to deemphasize anatomical detail, giving preference to the evaluation of vein function, which is considered most important. We believe that the low sensitivity and negative predictive value may be related to the operator’s motivation to identify clinically relevant variants rather than a reflection of the imaging modality itself. However, the comparison between preoperative duplex and intraoperative venography has shown that the preoperative study could perhaps be more focused on identifying any anatomical variants that may pose a technical challenge.
While the use of ionizing radiation in medical procedures is becoming more common, the decision to use adjunctive radiography must balance the individual risk of inducing a future cancer against the diagnostic and therapeutic benefits.21,22 Using the National Radiation Protection Board (NRPB-W14) methodology, we estimated the effective dose and resultant whole body exposure from each lower limb venogram to be 0.37 mSv, equating to a mean dose of 0.74 mSv per patient.21,23 This is less than the average naturally occurring background exposure experienced in the course of a year (2 mSv) and compares favorably to a number of common radiological procedures.21,22 DSV remains unchallenged in defining complex venous anatomy due to its clear, high-resolution image, and we believe that its use is individually justifiable where the potential health benefits of improving the success and durability of the procedure outweigh the very small risk.
The current practice of venous ablation is performed in both the outpatient setting and a hybrid operating theatre. The advantages of the hybrid theatre are the availability of a surgically sterile field and ready access to high-quality fixed C-arm imaging systems and DSV software. Venography uses few extra resources and takes little additional time. With ready access to venography and the advantage of achieving a complete treatment effect as demonstrated by this study, we now advocate the use of this imaging modality in complex venous anatomical configurations and have implemented it routinely in our practice.
Our study proved valuable in documenting the prevalence of venous anomalies, many of which may impede complete treatment of the incompetent venous segment as shown by the large proportion of patients requiring an endovascular maneuver to negotiate compromised vein segments. Furthermore, once x-ray guidance was used to negotiate these obstructions, it was possible to deliver additional radiofrequency energy or adhesive glue to effectively ablate large, dilated segments. Our most important finding was the large proportion of patients who were thought impossible to treat completely without adjunctive x-ray guidance. We believe that in the past these patients were left partially treated or converted to foam sclerotherapy, reducing durability. We recommend that it is these patients with complex anatomical variations in whom on-table fluoroscopy may be applied in addition to ultrasound to facilitate complete target ablation.
Limitations
The short follow-up in this study limited any conclusions related to long-term durability of procedures assisted by on-table venography. Mid- and long-term vein closure rates were not evaluated, as the study was focused on immediate treatment success and observations enhanced by venography. It is also true that leaving a portion of the target vein untreated or converting to foam sclerotherapy may not impact long-term outcome. While every effort was made to objectively quantify each venogram, it was impossible to avoid the subjective nature of an operator declaring a case “impossible.” However, each of these cases was carefully reviewed by multiple specialists and categorized to provide clarity. Finally, while preoperative ultrasound was found to be insensitive in the detection of venous anomalies, this result must be interpreted with caution. These duplex studies were not performed with the intent of detecting such anomalies but rather to ascertain valvular function.
Conclusion
This multicenter study demonstrated that venography is a useful adjunct in the treatment of superficial CVD, providing a more thorough evaluation of venous anomalies that may complicate the venous ablation procedure. Fluoroscopy- and venography-facilitated endovascular techniques were utilized in a large proportion of cases to facilitate complete saphenous vein ablative treatment.
Footnotes
Acknowledgements
The authors thank Mr Brent Rogers, CMLSO, MARPS, MACPSEM, Radiation and Laser Safety Officer, Prince of Wales Hospital, Sydney, Australia, for his assistance with the calculated estimates of whole body radiation exposure and effective dose from venography.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Ramon Varcoe is a consultant for Medtronic.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.