Central venous stenosis and occlusive disease (CVOD) is an increasingly frequent finding on venography. In the end-stage renal disease (ESRD) population, the majority of whom are initiated on catheter-based hemodialysis, the incidence ranges from 20 to 40, depending on the series.1,2 These patients may have adapted to their 50 intraluminal narrowing but more commonly exhibit an array of symptoms such as ipsilateral edema, pain, and access malfunction.3 The number, location, and duration of central venous catheters are all important factors in the development of CVOD.46 Moreover, the well-documented concurrent prevalence of chronic indwelling pacemaker and defibrillator wires in this population can further exacerbate CVOD symptoms.7 Although the precise mechanism of CVOD is unclear, a likely etiology is a combination of direct trauma, inflammation, and chronic endothelial injury.2,8,9


Intravascular ultrasound (IVUS) uses a 10 MHz to 40 MHz catheter-mounted probe to provide cross-sectional imaging within the vessel. This modality is commonly used for coronary interventions to assist with sizing, stent deployment, and assessments after percutaneous coronary interventions.1012 Recently, the Venogram vs IVUS for Diagnosing Iliac Vein Obstruction (VIDIO) trial demonstrated that IVUS showed a greater number of stenotic lesions and intraluminal diameter reductions in the iliofemoral system compared to traditional venography.13 Several studies have looked at using IVUS to assess CVOD in dialysis access patients, but consensus is lacking (Table 1).

Single-plane contrast venography and digital subtraction angiography (DSA) are the classic standard for evaluating CVOD, and the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (NKF KDOQI) guidelines recommend an endovascular-first approach to addressing COVD lesions.14 However, the accuracy of venography is limited in standard anterior-posterior projections, and multiplanar venographic studies are not common. Negln and Raju found that single-plane venography significantly underestimates the degree of stenosis in the iliofemoral system, and our unpublished institutional experience supports a similar assertion in hemodialysis patients with suspected CVOD (Figure 1).15 In the iliocaval system, stenosis is associated with decreased cross-sectional area and altered flow dynamics, but these outcomes have yet to be explored in the ESRD population.1618 We would expect a similar result, although it has been observed that vessel shape has a significant impact on cross-sectional area. Kabnick et al. showed that elliptical vessels had decreased cross-sectional areas compared to circular vessels with the same perimeter.19

Figure 1. 

Central venous stenosis on intravascular ultrasound (A) with reduced intraluminal diameter (yellow) and extrinsic compression (blue) compared to digital subtraction angiography (B).

Initial reports focused on using IVUS as an alternative modality for diagnosing CVOD. In 2008, Matthews and Thomas described using IVUS in an ESRD patient with severe left upper extremity edema and discomfort.20 The patient had a known anaphylactic reaction to iodinated contrast despite multiple pretreatment attempts. IVUS identified a focal area of stenosis confirmed with pullback pressures and amendable to balloon angioplasty.21,22 Carbon dioxide angiography had previously been the preferred modality for patients with severe reactions to iodinated contrast, but use in the thoracic cavity is relatively contraindicated due to concerns for air trapping and hemodynamic compromise.23

Tsai et al. later reported using IVUS in another ESRD patient with extremity edema to diagnose left brachiocephalic vein compression by the innominate artery.24 These findings were confirmed intraoperatively with venography prior to angioplasty and stenting. Pullback pressures and IVUS were able to identify residual stenosis missed on completion venography. Lin et al. reported similar findings in their series of 94 ESRD patients, wherein they observed that pullback pressure gradients were better predictors of long-term patency than venography; pressure gradients 5 mm Hg correlated with better outcomes.22 We do not routinely measure pressure gradients for CVOD in our practice but have anecdotally found that patients with gradients > 3 mm Hg may have significant stenoses. This echoes the experiences in the coronary and iliofemoral systems where IVUS was instrumental in guiding therapy.10,25

More recently, Graaf et al. compared DSA to intravascular ultrasound in a series of 12 patients with CVOD.26 IVUS showed residual areas of > 50 stenosis after angioplasty in six patients, whereas venography only identified those conditions in three patients. IVUS additionally identified trabeculae within the central vasculature that could not be visualized with standard venography. Although the authors could not establish the impact of these trabeculae, they postulated that such features may one day serve as additional criteria to guide intervention. Similarly, they note that the hyperechogenicity within the walls of stenotic vessels may represent fibrinous changes consistent with venous stenosis.


The growing body of data suggests that IVUS is a potentially useful adjunct in the diagnosis and management of CVOD in patients with ESRD. Although we cannot advocate for its routine use at this time, additional investigation in this patient population may lead to stronger recommendations.


  • Central venous stenosis and occlusive disease (CVOD) is frequently observed in end-stage renal disease (ESRD) patients.
  • Intravascular ultrasound (IVUS) may be better suited than traditional venography to identify intraluminal narrowing and pre-/post-intervention outcomes.
  • Additional study is warranted to better characterize the value of IVUS in the ESRD patient population.