Abd-Alraheem Hussein Ali ¹, Abd-Alraheem Fathy ², Ahmed Okasha Mohammed ¹, Mosab Ahmed Abdelsalam ¹

¹Department of Diagnostic and Interventional Radiology, Faculty of Medicine, Qena University, Qena, Egypt

²Department of Vascular Surgery, Faculty of Medicine, Qena University, Qena, Egypt

Abstract

Objective: This study aimed to evaluate the diagnostic utility of computed tomography venography (CTV) compared to Color Doppler Ultrasound (CDU) in patients with suspected secondary varicose veins.

Methods: This prospective diagnostic accuracy study included 100 consecutive patients with clinically suspected secondary lower limb varicosities evaluated between 2023 and 2025. Patients underwent detailed clinical examination, CDU, and 128-detector CTV. Evaluated parameters included the precise anatomical distribution of varicosities, central venous compression syndromes, and diagnostic agreement between the two modalities.

Results: The cohort comprised 51 (51%) males and 49 (49%) females (mean age: 36.4 (11.29) years). CTV definitively diagnosed May-Thurner syndrome in 31% of patients, pelvic congestion syndrome in 25%, lower limb deep venous thrombosis (DVT) in 12%, iliac venous thrombosis in 10%, Nutcracker syndrome in 5%, Klippel-Trenaunay syndrome in 4%, portal hypertension in 3%, Budd-Chiari syndrome, and inferior vena cava thrombosis in 2%. CDU failed to visualize the iliac veins in 60% of cases and was severely limited by patient obesity in 40%. For the detection of ovarian vein dilatation, CDU demonstrated a sensitivity of 93.3%, specificity of 80%, and an overall accuracy of 88% compared to CTV, with substantial agreement (κ = 0.73).

Conclusion: While CDU remains highly accurate for localized superficial and gonadal screening, it is insufficient for mapping central and peripheral venous pathologies. CTV is a method to be considered in patients with suspected secondary varicose veins, and in those whose causes cannot be determined by CDU.

Key words: Varicose veins, CT venography, color Doppler ultrasound, deep venous thrombosis, May-Thurner syndrome, pelvic congestion syndrome, iliac veins

Graphical abstract

Introduction

Traditional phlebography has gradually been replaced by more sophisticated non-invasive imaging modalities. One such technique that has shown promise is direct computed tomography venography (CTV). CTV enables accurate visualization and assessment of the precise location, size, and extent of venous lesions (1). A notable advantage of CTV over traditional Color Doppler Ultrasound (CDU) lies in its capability to effectively capture and detect venous abnormalities in both the pelvic and calf regions—areas that pose significant diagnostic challenges for CDU alone (2).

Current 2025-2026 guidelines from the Society for Cardiovascular Angiography and Intervention and the Society of Interventional Radiology recommendation are: CDU is the cornerstone for chronic venous disease (CVD); essential for assessing venous flow dynamics, locating reflux, and identifying anatomical abnormalities, transabdominal and transvaginal CDU is the preferred first-line imaging for pelvic venous disorders due to its ability to dynamically assess venous flow, reflux, and visualize periuterine/periovarian varices (3). 

CTV plays a crucial role in assessing varicose veins, a condition affecting 10%–40% of individuals aged 30–70 years. Determining the underlying cause of venous insufficiency is critical. In cases where varicose veins exhibit complex morphology or unusual manifestations, combining CDU with CTV aids in understanding the exact characteristics and causes of varicosities (4). By employing CTV, clinicians can acquire high-resolution images that reveal precise anatomical information (5), enabling a comprehensive evaluation of the vascular lesions and guiding targeted therapeutic strategies (6). While the diagnosis of deep vein thrombosis (DVT) and evaluation of iliac vein patency have historically been performed with CDU, helical CTV yields objective information about underlying anatomic abnormalities crucial for accurate diagnosis and evaluation of post-interventional thrombosis (7–9).

 This study aimed to evaluate the diagnostic utility of CTV compared to CDU in patients with suspected secondary varicose veins.

Methods

Study design and population

Study design and setting: This prospective diagnostic accuracy study was conducted on 100 patients admitted to Qena University Hospitals between 2023 and 2025.

Participants: The inclusion criteria comprised patients of all ages and both sexes with clinically suspected secondary varicose veins (e.g., pelvic congestion syndrome, DVT, Klippel-Trenaunay syndrome, May-Thurner syndrome). Morbidly obese patients in whom standard CDU were technically inadequate were also included. Patients with primary varicose veins, neoplastic/traumatic venous disease, prior venous surgery, or contraindications to CT contrast (severe renal impairment, pregnancy, allergy) were excluded. Patients were consecutively enrolled during the study period to reduce selection bias and better reflect real-world clinical practice.

Ethics:  The study was approved by the Institutional Ethics Committee. Written informed consent was obtained from all participants. The study was conducted in compliance with Helsinki 2024 agreement.

Baseline variables

 Detailed medical histories and clinical examinations were conducted. Baseline data included: age, body mass index (BMI), gender, parity (in female), lower limb varicose veins, pain, and edema, varicocele, vulvular varices, infertility, telangiectasia, hematuria, limb discrepancy, venous ulcers, and biles. Laboratory investigations included coagulation profiles, D-dimer, and renal function tests (serum creatinine via Jaffe’s Kinetic method).

Clinical examinations for varicose veins

Complaints as pain and edema were recorded.  Medical history and physical examination signs for lower limb varicose veins, pain, and edema, varicocele, vulvular varices were obtained.  Examinations were performed in standing and supine positions to map varicosities, pigmentation, and ulceration.

Imaging protocols for varicose veins

CT Venography

CTV was performed using a GE 128-detector row CT scanner (GE Healthcare, Chicago, IL, USA), (120 kV, 215 mA, pitch ≤ 1, 0.5 s rotation time, 0.6 mm slice thickness). Non-ionic iodinated contrast agent (iopromide 300 mg iodine/mL; Ultravist ® 300, Bayer Healthcare, Berlin, Germany) was injected at 4–5 mL/s (maximum 100 mL for adults) followed by a 50–60 mL saline flush. Bolus-tracking initiated scanning when femoral vein enhancement reached 120 HU. Scans extended from the diaphragm to the feet.

The evaluated CTV variables included venous compression syndromes (e.g., May-Thurner syndrome and Nutcracker syndrome), DVT, iliac and inferior vena cava (IVC) patency, pelvic venous congestion, ovarian vein dilatation, abdominal and pelvic collateral pathways, perforator vein mapping, venous reflux-related morphological changes, and associated anatomical variants or extrinsic compressive causes.

Color Doppler ultrasonography

CDU was conducted using a high-resolution ultrasound system (LOGIQ P7, GE Healthcare, Chicago, IL, USA) equipped with a 5-10 MHz linear transducer probe. Superficial, deep, and perforator veins were systematically assessed in standing (reflux) and supine (thrombosis) positions. Reflux was defined as retrograde flow >0.5 seconds.

CDU assessment included superficial and deep venous patency, lower limb deep vein thrombosis, saphenofemoral junction reflux, perforator incompetence, ovarian vein dilatation when visualized, pelvic venous congestion, and venous flow abnormalities.

Hemodynamically significant venous compression was defined as ≥50% reduction in luminar diameter compared to the adjacent normal venous segment.

Statistical analysis

Data were collected, coded, revised, and entered into the Statistical Package for the Social Sciences (SPSS) software program, version 27 (SPSS Inc., Chicago, IL, 2021). Qualitative variables are presented as numbers and percentages and compared using a McNemar test. Quantitative variables were tested for normality using the Kolmogorov–Smirnov and Shapiro–Wilk tests. Normally distributed numerical variables are reported as mean (SD) and range. Diagnostic performance (sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy) was assessed. Cohen’s Kappa was used for the agreement between CDU and CTV. Significance was set at p < 0.05.

Compliance with STROBE and STARD guidelines

The study was prepared in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines and the Standards for Reporting Diagnostic Accuracy Studies (STARD) recommendations.

Results

Baseline demographic and clinical characteristics

This prospective study included 100 patients evaluated for secondary lower limb varicose veins. Table 1 represents the demographics and clinical signs.

Clinically, lower limb varicosities were the predominant physical sign, present in 97% of the cohort. Gender-specific clinical manifestations included male varicocele, identified in 50.9% of male patients (26/51), and vulvar varices, observed in 7% of the total cohort.

A history of infertility was reported in 8% of patients, including both male and female individuals, predominantly female participants, due to the relation to underlying pelvic venous pathology. Symptomatically, patients presented with lower limb pain (36%), localized edema and redness (9%), telangiectasia (7%), and advanced venous ulcers (4%). Systemic and atypical manifestations were also recorded, including hematuria (7%), hematemesis (7%), biliary involvement (7%), known liver cirrhosis (7%), limb discrepancy (4%), and menorrhagia (3%).

CTV findings

Left common iliac vein compression (Fig. 1) was observed in 43% of patients, with the degree of stenosis ranging from 15% to 90% (mean: 55.1 (21.08)%).

Figure 1. (A) Axial contrast enhanced CTV of the abdomen showing left CIV compression between right CIA and the lumbar vertebral body, with luminal narrowing of the left CIV (orange arrow); (B) Axial contrast enhanced CTV showing dilated tortious congested pelvic veins; (C) 3D reconstruction showing the compressed left CIV and bilateral lower limb varicosities. (D) CDU  showing compression of the left CIV (white arrow). These findings are consistent with May-Thurner syndrome (non-thrombotic type)

CDU - color Doppler ultrasound, CIA - common iliac artery, CIV - common iliac vein, CTV – computed tomography venography

Overall venous compression within the cohort was classified as hemodynamically significant in 76% of cases, borderline in 1%, and non-significant in 23%. Renal vein compression  observed in 4% of patients was consistent with Nutcracker syndrome, characterized by compression of the left renal vein between the superior mesenteric artery and the abdominal aorta. Deep venous reflux was highly prevalent, identified in 65% of the cohort.

It was inferred based on indirect imaging features, including venous dilatation, collateral pathways, and associated pelvic congestion patterns.

Pelvic varices and congestion were detected in 30% of patients, while abdominal and pelvic wall varicosities were noted in 34%. Among female participants, substantial ovarian vein dilatation was present in 61.2% (30/49). Thrombotic events were accurately mapped, with lower limb DVT  present in 20% of patients and IVC thrombosis in 2%.

Detailed mapping of perforator veins demonstrated that leg and calf perforators were nearly ubiquitous (95%). Pelvic and gluteal perforators were identified in 15% of patients, followed by obturator (16%), inguinal (7%), lumbar (5%), lateral vein of Servelle (4%), sacral (3%), and pudendal (3%) perforators. Rare collateral pathways, including intrahepatic and azygos collaterals, were observed in 1% each.

Secondary anatomical variants and extrinsic factors identified via CTV included inguinal lymphadenopathy (9%), a bulky uterus (8%), uterine fibroids (12.2%), duplicated IVC (4%), left-sided IVC (4%), and retro-aortic renal vein (3%). One patient (1%) presented with an IVC JJ stent migration.

Final anatomical diagnoses

Based on comprehensive CTV findings, May-Thurner syndrome was the leading definitive diagnosis, present in 31% of the cohort (Table 2). This was followed by pelvic congestion syndrome (25%) (Fig. 2), lower limb venous thrombosis (12%), and iliac venous thrombosis (10%). Other diagnoses included Nutcracker syndrome (5%), primary varicose veins reclassified after ruling out secondary causes (6%), Klippel-Trenaunay syndrome (4%), portal hypertension (3%), IVC thrombosis (2%), and Budd-Chiari syndrome (2%).

Incidental extravascular findings were common and included splenomegaly (5%), ovarian cysts (5%), hydronephrosis (4%), gallbladder stones (4%), incidental liver cirrhosis (4%), paraesophageal varices (4%), ureteric stones (3%), left ovarian cystadenoma (3%), portal hypertension (2%), and enlarged congested prostate (2%). Prior trauma-related findings included right above-knee amputation (3%) and right iliac bone fracture via gunshot (2%).

Figure 2.  (A) 3D reconstruction of CTV showing multiple dilated tortious superficial and deep pelvic veins;  (B) Axial CTV  showing multiple dilated enhanced parametrial and para-ovarian pelvic plexus; (C) Coronal contrast enhanced CTV of the pelvis showing dilated tortious left anterior superficial pelvic veins;  (D) Note the dilated ovarian veins anterior to the psoas muscles on both sides (orange arrows).; (E) CDU  showing dilated left adnexal veins with reflux during Valsalva maneuver.  Findings are  consistent with pelvic congestion syndrome.

CDU – color Doppler ultrasound, CTV – computed tomography venography

CDU findings

 CDU successfully identified lower limb varicosities in 98% of patients. However, its detection of complex pathology was notably lower than CTV, observed as right iliac and deep pelvic thrombosis as shown in (Fig. 3), leg perforators were identified in 33%, saphenofemoral junction (SFJ) reflux in 32%, lower limb DVT in 9%, and compressed left common iliac vein in 18%. Central pelvic findings via CDU included deep pelvic venous congestion (14%), iliac/deep pelvic thrombosis (10%), and ovarian vein dilatation (9%) (Table 3).

Figure 3. (A) 3D reconstruction of CTV showing absent right CIV, IIV and EIV with multiple superficial and deep pelvic venous tortious collaterals; (B) Coronal CTV of the pelvic and both lower limbs veins; (C) Contrast enhanced CTV of the pelvis showing filling defect within the right EIV (white arrow) (note the difference of the contrast opacification of the normal left EIV). (D) CDU  showing left iliac DVT (yellow arrow). Findings consistent with extensive right deep pelvic venous thrombosis.

CDU – color Doppler ultrasound, CIV-common iliac vein, CTV – computed tomography venography, EIV- external iliac vein, IIV - internal iliac vein

Comparative diagnostic performance

When evaluating ovarian vein dilatation and pathological reflux (among 49 female patients) using CTV as the reference standard, CDU correctly identified 27 true positive and 16 true negative cases. This yielded a sensitivity of 93.3%, specificity of 80.0%, PPV of 87.5%, and NPV of 88.9%, with an overall accuracy of 88.0%. Agreement between the two modalities for this specific metric was substantial (κ = 0.73) (Table 4).

For the prediction of the final overall diagnosis, CDU correctly identified 82 true positives and 12 true negatives against the CTV gold standard. CDU demonstrated a sensitivity of 97.6%, specificity of 75.0%, PPV of 95.3%, and NPV of 85.7%, with an overall diagnostic accuracy of 94.0% (p < 0.001; McNemar test). The inter-modality agreement was substantial (κ = 0.74) (Table 5).

Discussion

This prospective study was conducted on 100 patients presenting with clinical suspicion of secondary lower limb varicose veins from Qena University Hospital between 2023 and 2025, and evaluated the diagnostic utility of CTV compared to CDU in patients with suspected secondary varicose veins. Among the 100 studied patients, 51% were males and the mean age was 36.4 (11.29) years. Among the 39 female patients, 83.7% were multiparous.

Our study demonstrated CDU has a good agreement with CTV findings with diagnostic accuracy in identifying secondary varicose veins of 94%. CDU however is highly accurate for superficial vein and gonadal varices, while low value in identification of deep venous structures. CTV offers precise visualization of deep anatomical structures, diagnosis of lower limb varicose veins, pelvic syndromes and central obstructive pathologies.

 The anatomy of the lower limb venous system is highly complex, making the assessment of venous insufficiency a diagnostic challenge. Varicose veins affect 10%–40% of the adult population and are rooted in venous valvular dysfunction (10). While CDU offers dynamic, real-time assessment (11), CTV adds significant value by identifying extrinsic compressions and complex congenital malformations often missed by ultrasound (4).

The cohort's mean BMI (25.7 kg/m²) and high multiparity rate reflect the cumulative effects of venous hypertension and structural vessel wall degradation. Obesity raises intra-abdominal pressure, and multiple pregnancies mechanically compress iliac veins, promoting permanent venous dilation (12). The slight male predominance (51%) aligns with occupational findings by Elumalai et al. (13) contrasting with the general clinical female predominance reported by Mulita et al. (14). The association between higher BMI and varicosities in the present study is strongly supported by Mok Ballew et al. (15) and Fusco et al., 2026 (16), reinforcing the heterogeneity of venous pathology across patient populations.

Clinically, the high incidence of male varicocele (50.9%), vulvar varices, and infertility points to upstream pathology like pelvic congestion syndrome or Nutcracker syndrome. Furthermore, portal hypertension manifestations (hematemesis, cirrhosis) increase systemic venous load (17). Doğan et al. (18) reported varicocele in 34.3% of patients with saphenous reflux, supporting elevated findings among study participants. Ulcer prevalence (4%) closely mirrored the 4.8% reported by Mulita et al. (14), though edema rates were lower in the current study. Clark and Taylor (19) confirm pelvic venous disorders affect approximately 15% of women, consistent with demographics in the current study.

Anatomically, left common iliac vein compression (43%) and reflux (65%) dominated the findings, driving a pathophysiology rooted in proximal venous outflow obstruction (20). Gavrilov (21) found pelvic congestion syndrome in 75.5% of women with pelvic varices, reinforcing the high pelvic disease burden observed in the current study. Wang et al. (22) noted left common iliac vein compression in 55.7% of patients, closely matching the 43% rate in the present study. Incompetent perforators were widespread (95%), a rate higher than the 58% reported by Zhang et al. (23), likely reflecting the combined imaging approach's higher sensitivity in this study.

Incidental findings such as splenomegaly and liver cirrhosis highlight systemic mechanisms elevating venous pressure (24).

The core diagnostic disparity lies in reflux detection: CTV identified reflux in 65% (deep/pelvic), while CDU found it in only 32% (predominantly superficial) (25). The prevalence of May-Thurner syndrome (31%) exactly matches findings by Heller et al. (26). Osman et al. (27) confirmed CT's high sensitivity (94.8%) for pelvic congestion syndrome. Conversely, baseline compression in unselected patients is much lower (14.7%), per Corrêa et al. (28). Badawy et al. (29) validate the heavy burden of iliac compression in symptomatic cohorts.

CDU failed to identify iliac veins in 60% of patients. This is driven by acoustic attenuation from deep adipose tissue and acoustic shadowing from bowel gas (30). Literature corroborates that CDU often underestimates iliac stenosis with high false-positive rates compared to cross-sectional modalities (31). In contrast, Clode et al. (32) found CDU sufficient for iliac stent evaluation in 99% of their cohort despite obesity, highlighting a divergence depending on the specific post-operative vs. diagnostic setting. We emphasize that CTV may be most appropriately utilized in selected patients, particularly in cases where CDU findings are inconclusive or when more detailed anatomical evaluation is required.   

The primary strength is the prospective, dual-modality design. Limitations include the single-center setting and the inherent risks of ionizing radiation and contrast administration associated with CTV.

The diagnostic utility of CDU was substantially limited by patient habitus and deep anatomical structures. Obesity limited the sonographic window in 40% of cases. Consequently, CDU failed to identify the internal iliac veins in 60% of patients, the renal veins in 30%, and the ovarian veins in 32.65% of female patients.

Conclusion

Computed tomography venography provides indispensable anatomical mapping in the evaluation of secondary lower limb varicose veins, definitively identifying central obstructive pathologies such as May-Thurner and pelvic congestion syndromes. While Color Doppler Ultrasonography remains highly accurate for localized superficial and gonadal screening, its high failure rate in visualizing deep pelvic structures limits its standalone utility. A larger-scale, multicenter randomized trials are recommended for more informative findings.

Ethics:  Written informed consent was obtained from all participants. This study was approved by the Institutional Ethics Committee of Qena University Hospitals and conducted in accordance with the Declaration of Helsinki (2024).

Peer-review: External and internal

Conflict of interest: None to declare

Authorship: Study concept and design: all authors;  Data acquisition: A.-A.H.A., A.O.M.,  Data analysis: M.A.A., manuscript drafting: M.A.A.; Critical revision and approval final version  for publication: all authors. Thus all authors  fulfilled authorship criteria.

Acknowledgements and Funding:  None to declare.

Statement on A.I.-assisted technologies use: No artificial intelligence tools were used for image generation, data analysis, or scientific decision-making. Any language assistance tools used during manuscript preparation did not influence the scientific content of the study.

Data and material availability: De-identified participant data supporting the findings of this study are available from the corresponding author upon reasonable academic request. Any sharing should in frame of fair use an collaboration with acknowledgement of source