Computed Tomography Venography

CT venography (CTV) has a long history; the first diagnoses of caval and iliofemoral thrombi on CT were made in the 1980s. CT has some inherent advantages over MRI for venous imaging. Among these are less dependence on the technologist, extended anatomic coverage (potentially including the lower extremities as well as the abdomen and pelvis) in a single continuous data set, and generally very high spatial resolution. Three-dimensional (3-D) reconstructions can be performed more easily than with MRI, and they often include osseous structures for easy anatomic reference.

Although nephrotoxicity of iodinated contrast is a potential concern in the setting of renal insufficiency, there is generally no contraindication to contrast administration in patients undergoing chronic hemodialysis. Iodinated contrast is generally considered safer for dialysis patients than gadolinium contrast for MRI because of the risk of nephrogenic systemic fibrosis with gadolinium agents in the setting of severe kidney disease. CT is also the method of choice in patients who have devices that are not compatible with MRI, including pacemakers and implanted defibrillator devices, deep brain or spinal stimulators, and medication pumps.

Potential disadvantages of CT with respect to MRI include exposure of the patient to ionizing radiation (which may be compounded by subsequent fluoroscopic procedures and then follow-up CT examinations), higher incidence of allergic reaction to iodinated contrast than to gadolinium, lower vessel-to-background contrast, and low intravascular concentration of iodine (particularly in indirect venography, described later), leading to a high incidence of nondiagnostic studies. In addition, unlike MRI, CT does not afford the ability to adequately image veins without contrast or to quantify magnitude or direction of venous flow.

CTV can be performed through direct or indirect approaches, with contrast injected distal to the anatomic region of interest in the former case and imaged as the contrast travels antegrade toward the heart. Alternatively, contrast can be injected at a remote location (such as an unrelated peripheral arm vein) and imaged in a recirculation phase in the latter case. When performed optimally, direct CTV generally tends to generate superior images, but timing of the injection with respect to the scan can be difficult. Standard iodinated contrast preparations must be diluted to a quarter or third of their original concentration with saline to limit streak artifact, and if bilateral upper or lower extremity imaging is needed, IV access must be established separately and at least two technologists or nurses are required to inject at similar rates in each extremity simultaneously. Timing of venous contrast transit can be quite variable (and asymmetric) and multiphasic scanning is often required, increasing the patient’s radiation exposure.

Although indirect CTV is less technically challenging to perform, the contrast bolus is essentially diluted by the patient’s entire blood volume. This necessitates administration of a higher volume of contrast (often 200 mL in our practice) and nevertheless results in an 11% rate of nondiagnostic studies. However, the technique still can be quite powerful and does show adjacent visceral anatomy as well (Figure 1).

One potential advantage of CTV is that it can be combined with CT pulmonary angiography to provide a single examination (Figure 2) for pulmonary embolus (PE) and deep vein thrombosis (DVT). This is accomplished by imaging the pelvis and thighs 2 to 4 minutes after injection of the contrast bolus for pulmonary CT angiography (CTA), allowing venous recirculation and opacification of lower extremity veins. This has the advantage of rapid acquisition and use of a single contrast bolus.

A potential negative aspect of this approach, however, is the additional radiation exposure to patients, many of whom are quite young and likely to accumulate a high lifetime radiation exposure. When imaging the pelvis and lower extremities, exposure of radiosensitive gonadal tissue in these patients cannot be avoided. In addition, a large meta-analysis of 24 studies with 17,373 patients showed that adding delayed CTV to CT pulmonary angiography resulted in only a 3% increase in detection rate of thromboembolic disease. This study also demonstrated that using Medicare reimbursement rates, significant cost savings ($2,013 per incremental DVT) can be realized by performing duplex ultrasound on patients with a negative pulmonary CTA rather than performing delayed CTV, with the additional advantage of decreased radiation exposure.

Although imaging venous stent patency is challenging with CT and MRI, CT is probably the better option for most patient with most stents. This is mainly a result of metal artifacts from the stents causing more artifacts on MRI than on CTA, particularly in small vessels. CT should be considered the test of choice for patients with contraindications to MRI because of implanted hardware, as well as for patients on hemodialysis, particularly if the presence or absence of distal venous thrombi will affect treatment. CT should probably be the first test when investigating venous stent patency if ultrasound is equivocal or technically limited.