Radiography

Inspection of an abdominal radiograph is valuable to note the presence and approximate location of stents, surgical clips, or other material that might interfere with subsequent imaging. This is especially true for MRA, in which local loss of signal because of clip artifact is a well-recognized pitfall.

Ultrasound

Like native renal artery stenosis, the diagnosis of TRAS is made with ultrasound by demonstrating a focal area of aliasing on color Doppler imaging from turbulence caused by the stenosis and an elevated peak systolic velocity through the stenosis more than 2.5 m/sec on spectral Doppler imaging (Fig. 112-2).⁵ Because of the proximity of the allograft to the body wall in the anterior pelvis, the main artery can often be visualized in its entirety, aiding the chances of confirming the diagnosis. Power Doppler, with its high sensitivity to flow in all directions, is helpful to localize and map the artery for further evaluation with color Doppler and placement of the spectral Doppler gate. However, transplant arteries are typically much more tortuous than native renal arteries. This makes the setting of accurate angle correction on spectral Doppler more difficult and sometimes almost impossible. Furthermore, a range of elevated velocities has been shown in transplant arteries, even when stenosis is not suspected or proven not to be present, particularly if there is tortuosity.⁶ When there is no significant curvature, an elevated velocity may reflect elevated velocity in the external iliac artery. In these cases, a ratio of velocities in the main transplant artery and external iliac artery less than 1.8 is unlikely to be stenosis.

FIGURE 112-2 High-velocity flow with spectral broadening indicating severe stenosis of transplant artery end-to-side anastomosis to the external iliac artery.

As in native renal stenosis, the characteristic parvus-tardus waveform of the intrarenal arteries downstream from a stenosis can support the diagnosis. This is reflected in spectral broadening of the arterial waveform, retarded acceleration less than 1.5 m/sec², and increased acceleration time more than 0.08 to 0.10 second (Fig. 112-3) However, these findings have been shown to be less specific than peak systolic velocity and should be used to support the diagnosis based on peak systolic velocity.⁷ Resistive and pulsatility indices can also be lowered in TRAS but are only nonspecific indicators of graft dysfunction.

FIGURE 112-3 Low-resistance flow and tardus-parvus waveform with delayed systolic peak detected on Doppler ultrasound of intrarenal arcuate artery.

Computed Tomography

This is not commonly used to assess transplant renal artery stenosis due to the nephrotoxicity of iodinated contrast.

Magnetic Resonance Imaging

MRI can demonstrate TRAS with a high degree of accuracy. In multiple series, gadolinium (Gd)-enhanced three-dimensional MRA has been shown to have a sensitivity of 67% to100% and a specificity of 75% to 100% for hemodynamically significant stenosis (Table 112-1). Although technically challenging on older scanners, the examinations have become routine with state of the art 1.5-T scanners.

Gd-enhanced MRA is performed with a three-dimensional spoiled gradient echo pulse sequence. With a fast repetition time (TR), multiple three-dimensional acquisitions (i.e., multiphase), each taken during a single breath-hold, can be obtained in approximately 2 minutes. Acquiring the first set before the administration of gadolinium allows for image mask subtraction of the precontrast image data set from postcontrast data sets, which can then be processed on a dedicated three-dimensional workstation for improved imaging using a variety of postprocessing visualization tools, such as multiplanar reformation, maximum intensity projection (MIP), and volume rendering. Careful timing of the contrast bolus to synchronize peak arterial Gd concentration with central k-space data acquisition is essential and is usually achieved by automated triggering or MR fluoroscopy sequences. Using phased-array surface coil arrays with parallel imaging helps shorten the duration of center of k-space to minimize artifacts and maximize resolution and anatomic coverage for a comfortable breath-hold time. Twofold acceleration is possible with eight coils and four- or fivefold acceleration works well with 32 coils. High resolution is necessary to resolve the renal artery adequately, with slice thickness preferably less than 3 mm zero interpolation down to less than 1.5 mm, and in-plane resolution less than 2 mm. The Gd dose should be kept at 0.1 mmol/kg or less to minimize risk of nephrogenic systemic fibrosis (NSF) in case the glomerular filtration rate (GFR) is less than 30 mL/min. Promising new noncontrast renal MRA techniques are also emerging but are not yet validated.

Review of source images, as well as review of three-dimensional postprocessed projectional views, is necessary to make the appropriate diagnosis. MIPs and multiplanar reformations using variable thicknesses allow for direct visualization through the lumen of the transplant renal artery, the anastomosis, the iliac arteries, and usually the first-order branch arteries (Fig. 112-4). It is the ability to manipulate the volumetric data of MRA that makes it so effective. This is especially true in tortuous vessels, which are typical in renal transplants. Shaded surface volume renderings can be especially helpful to visualize the anatomy in the case of tortuous vessels.

FIGURE 112-4 Severe stenosis at transplant artery end-to-side anastomosis onto external iliac artery. A, Anteroposterior (AP) MIP of entire imaging volume. B, Oblique subvolume MIP oriented for optimal display of transplant renal artery.

An effective complement to Gd-enhanced MRA is three-dimensional phase contrast MRA. This sequence achieves the MR signal, and therefore contrast, from protons moving across flow-encoding gradients. The signal is velocity-encoded at a certain upper threshold so that superfast flow at severe stenoses nulls the signal. Although the sequence acquisition time is long (typically 5 to 8 minutes), phase contrast MRA can be performed on a small field of view, either in the area of the transplant renal artery or any area of suspicion during imaging. Confirming suspected stenoses by demonstrating loss of signal on phase contrast helps eliminate false-positives. The combination of Gd-enhanced MRA and phase contrast MRA has been shown to increase specificity from 88% to 100%.⁸

The limit of MRA resolution for stenosis is typically at the level of the segmental first-order branch vessels. DSA is superior in its ability to detect segmental artery stenosis. Furthermore, intrarenal arteries are typically obscured by enhancement of the renal parenchyma with MRA. However, because MRA is an anatomic image that directly displays the vessels, similar to digital subtraction angiography (DSA), it offers many advantages over the detection of secondary findings of TRAS with Doppler ultrasound (US). For example, kinked or tortuous arteries that are technically challenging or impossible with Doppler US can be evaluated with the volumetric data sets of MRA. Iliac artery stenosis proximal to the transplant artery anastomosis is an uncommon cause of TRAS (pseudo-TRAS) that is much more easily identified with MRA (Fig. 112-5). Because of the shortage of grafts, en bloc transplantation of two kidneys of limited function such as from the cadavers of older patients or two small kidneys from pediatric cadavers, has gained acceptance (see Fig. 112-1C). In these cases, not only is the anatomy more complicated, but secondary findings of velocity or ratios of velocities have not been proven to be reliable indicators of TRAS. Direct anatomic imaging also allows for the detection of other pathologies that may be the cause of patient symptoms, such as infarctions, artery or vein thrombosis, urinary collecting system complications, or a perinephric collection. It can also be helpful for preoperative planning, even when DSA is already indicated.

FIGURE 112-5 Transplant artery anastomosis is widely patent but there is external iliac artery anastomosis proximal to the transplant artery (arrow). This is a site where the artery was injured from a clamp during transplantation surgery.

The major limitation of MRA in the detection of TRAS is metal artifact from surgical clips. This can often be identified as a focus of complete signal dropout with an adjacent bright focus of displaced signal. In cases of suspected recurrence in treated stenoses, the presence of an endovascular stent usually limits the ability to make the diagnosis with MRA because of artifact. MRA is effective only when the stented segment of the artery is shown to be widely patent, because loss of signal caused by artifact cannot be discriminated from a recurrent stenosis.

A recent concern in the use of Gd-enhanced MRI is NSF. This is very rare, but is a debilitating and largely untreatable disease thought to be related to the deposition of gadolinium in the skin after dissociation from the chelated form used in MRI contrast agents. The Gd dose should be kept at 0.1 mmol/kg or less to minimize the risk of NSF in case the GFR is low (see earlier). Other risks for NSF include a concomitant inflammatory state, such as recent surgery or deep venous thrombosis, and the elevated phosphate levels of patients in renal failure. Although the large majority of cases have been reported in patients on dialysis, a GFR less than 30 mL/min is a relative contraindication for the administration of Gd chelate contrast agents. When the need to determine the diagnosis or exclusion of TRAS outweighs the risk of NSF, steps should be taken to ensure that the patient is well hydrated, that acute inflammatory states have passed before MRA is performed, that low-dose or noncontrast protocols are attempted first and, most importantly, that patients on dialysis are dialyzed immediately after receiving gadolinium (within 24 hours).

Nuclear Medicine and Positron Emission Tomography

Although a decline in renal function after the administration of an ACE inhibitor can indicate TRAS, an ACE inhibitor challenge during scintigraphy is relatively contraindicated because of reports of ACE inhibitor–induced acute renal failure, in some cases irreversible. Additionally, the specificity is less than in US, MRA, and DSA.⁹

Angiography

Catheter angiography with iodinated contrast is the gold standard for the diagnosis of transplant renal artery stenosis. It offers the highest resolution available for visualizing stenoses, including stenoses in branch renal arteries. Additionally, hemodynamic significance can be directly tested after crossing a stenosis with a catheter. A pressure gradient across a focal or a segmental stenosis more than either 10% or 10 mm Hg indicates hemodynamic significance. As an invasive test, however, catheter angiography has the highest associated risk, with mild expected morbidity and occasional severe unexpected morbidity. It is also the most expensive.

Ultrasound

Color-flow Doppler and point spectral Doppler imaging are the mainstays of diagnosing AVF and PSA with ultrasound. Although real-time ultrasound may show an anechoic focus corresponding to an enlarged vein or a PSA sac, this is usually only the case with large or extrarenal lesions. The most common finding is a focal area of increased color saturation caused by tissue vibration around the AVF.¹⁴ Once this area has been localized, spectral Doppler investigation of the feeding artery usually shows increased velocity, with increased diastolic flow and spectral broadening (Fig. 112-6). The draining vein almost always shows increased pulsatility because of arterialization, but this finding by itself can be seen in other problems, such as right heart failure. The arterial and venous spectral findings should be compared with branch vessels in other segments of the transplant kidney to increase specificity. PSAs appear as an area of swirling, turbulent flow on color Doppler. There is biphasic to and fro flow through the neck of the cavity on spectral Doppler.

FIGURE 112-6 A, Ultrasound shows an AVF (arrow) following renal biopsy. Note the spectral broadening from the high velocity flow. B, Doppler ultrasound in another patient with AVF following renal biopsy.

Magnetic Resonance Angiography

Contrast-enhanced three-dimensional MRA can demonstrate an AVF or PSA anatomically. The hallmark of an AVF is an early draining vein on the arterial phase of imaging. With the use and manipulation of MIPs, the actual fistulous connection can often be visualized. Similarly, a PSA can be demonstrated, including the size and dimensions of the neck and any thrombosed portions of the lesion (Fig. 112-7). Time-resolved three-dimensional MRA provides multiple sequential data sets that can provide visualization of early arterial filling of these vascular lesions.

FIGURE 112-7 Three-dimensional Gd MRA shows a severe stenosis at the end-to-end renal transplant anastomosis to right internal iliac artery. Note also a small pseudoaneurysm (bottom arrow) just distal to the anastomotic stenosis (top arrows).

Angiography

Catheter angiography using digital subtraction can image AVFs or PSAs anatomically and is performed during treatment as described below.

Ultrasound

Graft artery thrombosis is diagnosed by an absence of flow in the main artery and vein. Graft vein thrombosis shows echogenic thrombus in the occluded or partially flowing vein. A swollen hypoechoic kidney is usually an associated finding in venous thrombosis. An elevated arterial resistive index is usually also seen because of the lack of collateral outflow in transplant kidneys, but this is nonspecific, because it is also seen in rejection or acute tubular necrosis. Reversal of diastolic arterial flow on spectral Doppler, once thought to be diagnostic of venous thrombosis, is also nonspecific.

Magnetic Resonance Angiography

Renal artery thrombosis manifests as nonfilling or nonvisualization of the artery in the arterial phase of contrast-enhanced (CE) MRA, followed by global nonperfusion of the graft in the parenchymal or venous phases. Segmental infarctions caused by segmental artery thrombosis can also be exquisitely demonstrated on MRA. The finding of loss of corticomedullary differentiation on T1-weighted images has also been described, but is a nonspecific finding that is also seen in rejection and cyclosporine toxicity, for which biopsy remains necessary for definitive diagnosis.

A lack of filling of the vein on delayed CE-MRA phases, sometimes with a thin rim of enhancement probably representing the wall of the vein, is diagnostic of renal vein thrombosis. This can be more difficult to demonstrate as the gadolinium bolus is diluted. Careful localizing of the vessels prior to contrast administration, using a bright blood imaging technique, notably steady-state free precession (SSFP) imaging, is helpful to ensure inclusion of vascular structures in the field of view. There may be a loss of the normal flow void on T2-weighted images. A secondary finding of renal vein thrombosis is a delayed and persistent nephrogram. Diagnosing more proximate thrombosis in the iliac vein or inferior vena cava is another advantage of CE-MRA.

Computed Tomography

Multidetector CT with a rapid contrast bolus and imaging in arterial, parenchymal, and excretory phases provides excellent delineation of the renal vasculature, including accessory vessels, with almost 100% sensitivity. Thin slice reformations are used to allow for improved three-dimensional postprocessing. MIP is the most superior and time-efficient format to evaluate the vasculature, but the source data should be reviewed because it is the most accurate for small accessory vessels. Renal vein variants, such as retroaortic or circumaortic veins, and early arterial branching (within 2 cm of the origin of the artery) should be noted because they will change the surgical approach. A precontrast set of images is usually obtained to exclude occult renal calculi, a relative contraindication to donation. The parenchyma should be evaluated for masses, cysts, or scarring. Imaging during the excretory phase is usually performed throughout the entire collecting system to screen for occult urothelial lesions or variants. For example, completely duplicated ureters will usually lead to harvesting of the contralateral kidney but partially duplicated ureters may not.

Magnetic Resonance Angiography

Dynamic CE-MRA is almost as accurate as CTA for detection of the main renal vessels and their variations and is comparable for detection of accessory arteries larger than 1 to 2 mm. As with CTA, evaluation of MIPs and source images should be performed. An excretory phase of imaging should be performed (MR urography) and can detect ureteral duplications. MRI is not sensitive to small urothelial lesions, unlike CT urography; however, these lesions are rare in healthy renal donors. The sensitivity of MRI for urothelial lesions can be improved by injecting furosemide (e.g., 10 mg dose) after the Gd chelate contrast agent and performing a T1-weighted three-dimensional spoiled gradient-recalled echo (SPGR) sequence to image the distended collecting system. MRI is not sensitive to small renal calculi. The advantage of MRA is the relative safety of Gd chelate contrast and its lack of associated ionizing radiation exposure, which a similar multiphase CT urogram would require. Thus, MRA may be the modality of choice in patients with a contraindication to iodinated contrast or younger patients who have greater long-term risk of malignancy related to lifetime cumulative ionizing radiation exposure.

Angiography

DSA is more sensitive to fibromuscular dysplasia (FMD) than CTA or MRA, specifically mild disease in the proximal arterial segment and mild or moderate disease in first-order branches. For the general population, the risk of rendering the donor to a single kidney afflicted with FMD is thought to be less than the risks associated with an invasive procedure. DSA could be considered if the potential donor is at greater than average risk for FMD.

Ultrasound

Color and spectral Doppler ultrasound can usually identify the iliac vessels, donor Y graft, and main graft vessels when the pancreas transplant is visualized.²⁵ However, because of its intraperitoneal location, the graft may be obscured by bowel. As in renal transplants, power Doppler is useful to identify vessels, but diagnosis of venous or arterial thrombosis requires demonstration of a lack of flow on spectral Doppler. A lack of venous flow in combination with reversed or severely dampened diastolic arterial flow has been shown to be highly accurate for venous thrombosis.²⁶ Similarly, arteriovenous fistulas and pseudoaneurysms appear on color Doppler as tissue vibration surrounding a high-flow artery and draining vein or swirling flow within a saccular outpouching from the artery, respectively, but should be confirmed with typical spectral Doppler waveforms.

Computed Tomography

CT with iodinated contrast can demonstrate vascular complications of pancreatic transplantation but should be avoided when a transplanted kidney is also present.

Magnetic Resonance Angiography

Because of the complex vascular anatomy, CE-MRA, with three-dimensional volumetric data sets, is the modality of choice to evaluate the vascular complications of pancreas transplantation. In a recent study using state of the art high-resolution three-dimensional CE-MRA, all surgically or angiographically proven vascular complications, including venous and arterial thrombosis, arteriovenous fistula, pseudoaneurysm, and stenoses, were diagnosed.²⁷ Occasional false-positives, especially overgrading stenoses, still occur. Direct anatomic visualization of lesions, as in renal transplants, is accomplished with manipulation of MIPs and correlation with the source images. All main vessels and 85% of first-order branches can be visualized. Additional MRI sequences, including parenchymal enhancement, can be used to rule out other causes of graft dysfunction, such as enzymatic leak or peritransplant collections and to assess for possible infarction or infection of the graft.

KEY POINTS

Venous thrombosis is the most common vascular complication of pancreatic transplantation and usually occurs within the first week.

Ultrasound can often visualize the graft and its vasculature quickly and portably.

MRI-MRA is the modality of choice on which to base treatment decisions because of its high accuracy for visualizing the complex pancreas transplant anatomy and its facility for evaluating for other causes of graft dysfunction.