Expanding contrast options widen clinical utility of MRA

February 11, 2006

MR physics dictates that the faster the blood flow, the better the likelihood of quality vessel imaging. Attaining this goal in clinical practice is clearly difficult, given the need for patients to lie still on an MR table for up to 45 minutes. The problem of sluggish blood flow is further compounded in patients with circulatory problems. Yet these are the very individuals who may require MR angiography the most.

MR physics dictates that the faster the blood flow, the better the likelihood of quality vessel imaging. Attaining this goal in clinical practice is clearly difficult, given the need for patients to lie still on an MR table for up to 45 minutes. The problem of sluggish blood flow is further compounded in patients with circulatory problems. Yet these are the very individuals who may require MR angiography the most.

Most MRA examinations consequently require addition of a contrast agent to produce diagnostic-quality images (see accompanying article). Agents used for MRA have a far better safety profile than those employed for CTA. The majority of MR agents are made from gadolinium, which in itself is highly toxic. When bound to a chelate, however, the toxicity of Gd is removed.

"With iodinated CT contrast agents, you have to worry about prehydration of patients, nephrotoxicity, and cardiotoxicity. There is also the potential, albeit rare, for lactic acidosis in diabetics who are taking Glucophage. None of these issues exist with Gd-based MR contrast agents. They are virtually completely safe," said Dr. James Meaney, a radiologist at St. James's Hospital in Dublin.

Standard contrast-enhanced MRA involves intravenous injection of a bolus, which is then captured in a snapshot image once it reaches the area of interest. Timing the start of imaging is critical. Too soon, and the MR signal may be too weak. Too late, and visualization will be hampered by venous enhancement.

The tricky issue of bolus timing can be tackled in one of two ways. Many radiologists like to inject a test bolus to assess the contrast kinetics. This has the advantage of tailoring timing to each patient's individual physiology. An alternative approach is to purchase a commercial bolus-tracking software package that can automate the start of 3D imaging once contrast concentration in the field-of-view has reached a certain level.

Opinions differ as to which is the better approach. Dr. Mathias Goyen, a radiologist at the University Medical Center Hamburg-Eppendorf in Germany, prefers the traditional test bolus method for most of his MRA examinations. Some criticize the method as being too time-consuming. Others appreciate the fact that the method keeps radiologists, and not software, in control.

"For example, to image the renal arteries, we first inject 1 or 2 mL of contrast into an antecubital vein and then image the aorta at the level of the renal arteries every second. When we see peak enhancement, we note the time," Goyen said. "We then know how long after injecting the bolus we should start free acquisition to really catch the contrast in the arterial phase."

The required volume for the main bolus injection will depend on the patient being imaged. Dosages typically vary from 0.1 to 0.2 mmol/kg body weight. This translates into about 20 to 30 mL of contrast for a 75-kg patient. The maximum recommended dose is 0.3 mmol/kg body weight.

In addition to perfecting the timing, Goyen also employs a number of far simpler strategies to curb problems with venous overlay. For instance, application of midfemoral compression with a tourniquet or blood pressure cuff reduces the occurrence of venous overlay in the infrapopliteal arteries, he said.

Meaney, conversely, is a firm believer in automated detection of contrast arrival in the region of interest. Bolus-tracking packages are expensive. But once installed, there are no disadvantages to using the automated functionality, and several advantages, he said.

"It's the idiot-proof method," he said. "You watch in real-time for contrast arriving in the ROI on the screen, and then you hit 'go.' Because the sequence is routinely set up with centric/elliptical centric phase encoding, you are virtually guaranteed selective arterial phase images without confounding venous enhancement every time."

Dr. Jonathan Gillard, a neuroradiologist at Cambridge University Hospitals Foundation Trust, U.K., agrees that timing is critical for standard steady-state MRA examinations. Having tried a variety of methods, he has concluded that the best option is to find a good technologist with a cool head, who will not trigger the start of imaging too quickly.

Gillard now performs contrast-enhanced MRA for all carotid artery imaging. The technique is just as good as conventional angiography, it is cost-effective, and patients prefer it, he said. But many institutions are still performing unenhanced carotid MRA or sticking with conventional imaging strategies.

"If everyone switched now to contrast-enhanced MRA for carotid disease, it would be a great step forward," he said.

ADDED DYNAMISM

One additional bonus of MRA over CTA is the potential for dynamic imaging. Protocols for time-resolved imaging involve the rapid acquisition of multiple frames over time, in a similar manner to digital subtraction angiography. Such repeated imaging would be unworkable for CT, owing to the high radiation burden from multiple scans.

Time-resolved MRA permits lower doses of contrast to be administered, according to Dr. James Carr, director of cardiovascular imaging at Northwestern University Medical School in Chicago. Injection of a 20 mL bolus makes no sense if the total acquisition time is just 500 msec per frame. Diagnostic-quality images should be achievable with just a few mL of contrast.

"That's the beauty of time-resolved MRA. First of all, it allows you to get dynamic information about the blood vessels, but it also allows you to use less contrast," he said.

Time-resolved techniques are best suited to applications that are not dependent on good breath-holding, Carr said. When imaging the renal arteries, for instance, it can take 15 to 20 seconds for injected contrast to reach the required area. Asking a patient to hold their breath much longer than this in order to obtain an imaging series is impractical. This is not an issue when imaging extremities, such as the hands or feet.

Carr additionally recommends time-resolved techniques for pulmonary MRA. Contrast arrives very quickly in the chest, so the breath-holding time is not unreasonably long. He also combines steady-state and time-resolved techniques in a single MRA examination. When imaging the abdomen, for example, time-resolved MRA can be used first as a substitute timing bolus, with the advantage of providing dynamic information. This can be followed by a second, larger contrast injection to get a high-resolution snapshot.

"Because we now have all these different ways of doing MRA, you can tailor the examination depending on what you are actually looking for. It is much more powerful than it was five or six years ago," Carr said.

Dr. Stefan Schoenberg, chief of MRI at University Hospitals Grosshadern in Munich, often combines steady-state MRA with a time-resolved scan. A typical MRA exam at the Munich hospital starts with a time-resolved perfusion measurement, aiming for one- or two-second temporal resolution and 2 to 3 mm isotropic resolution. This is performed with the assistance of either 0.5 M contrast with the highest possible T1 relaxivity, or a higher concentration 1 M agent. Assessment of kinetic flow is then followed by steady-state imaging to generate 1 mm isotropic data sets.

Time-resolved techniques can be of particular benefit in the distal lower extremities, where bolus tracking can be difficult, said Dr. Thomas Grist, a professor of radiology at the University of Wisconsin, Madison. Precision timing of contrast arrival can also be problematic in these vessels if an occlusion is present on one side and not the other. Time-resolved techniques can reveal unusual filling patterns in patients with severe disease. Such details are critical when planning appropriate treatment.

The optimal approach to peripheral artery imaging is generally dictated by the patient's condition, according to Dr. Tim Leiner, a radiologist at Maastricht University Hospital in the Netherlands. Patients presenting with relatively mild arterial disease, or intermittent claudication, typically undergo MRA to identify any short occlusions or stenoses that can be dilated easily. In patients with rest pain, or critical ischemia, MRA needs to identify the patency of upper and lower leg vessels to assess the viability of surgical bypass. The chosen protocol should cater to the different clinical questions in each case.

For patients with critical ischemia, Leiner prefers to inject two separate contrast boluses and perform two separate acquisitions. This makes it easier to obtain good lower leg images that are free from venous enhancement, he said. The initial acquisition is taken of the lower legs and typically requires 20 mL of 0.5 M contrast. A moving table acquisition is then performed with a second contrast bolus to image the aortoiliac and upper leg arteries. The total dose should not exceed 0.3 mmol/kg body weight.

"It takes a bit more time, because you have two injections and two acquisitions, but the image quality that you usually achieve is of such high quality that the evaluation becomes much easier," he said.

BLOOD POOL BENEFITS?

Lengthy examinations can become an issue in institutions where MR table time is in great demand. The emergence of new blood pool contrast agents that remain in the vasculature for one or two hours may help, Leiner said. That means that imaging need not be performed immediately after contrast injection, and agents can even be administered before patients enter the scanning room.

The use of blood pool agents may also permit repeat imaging if mistakes are made in an initial acquisition, said Dr. Martin Prince, chief of MRI at the Presbyterian Hospital and Weill Cornell Medical Center in New York City. Repeat contrast-enhanced MRA should be possible with a single injection of an intravascular agent. The initial dose of contrast will probably be lower as well, given the greater T1 relaxivity of blood pool agents. Alternatively, the same dose could be used to ramp up the signal-to-noise ratio to get higher resolution.

"The full clinical utility of these agents has yet to be discovered," he said. "But let's take an example. Imagine you are worried about renal artery disease in a patient with diabetes. If you're injecting a blood pool agent, there is no reason why you couldn't image the entire vasculature. So while they are in the scanner, you could take a few extra minutes and check for peripheral vascular disease, look for signs of any previous myocardial infarction, and assess changes in the carotid arteries."

Blood pool agents may assist in areas with complex venous anatomy, where veins and arteries are close together, Grist said. This is frequently the case in the lower legs, where paired veins are often immediately adjacent to arteries. The close proximity of carotid arteries and jugular veins can also make it difficult to distinguish the individual vessels. Combination of time-resolved imaging with a prolonged steady-state study should allow the arteries and veins to be segmented.

The key strength of intravascular agents will be in multiple organs and whole-body MRA, according to Schoenberg. High-resolution steady-state MRA is already sufficiently accurate for imaging single territories. As medicine and radiology move toward disease-specific treatments, however, contrast agents that can broaden the coverage of an MRA examination will come into their own.

The resistance to leakage exhibited by blood pool agents means they may be ideally suited to MR perfusion studies in the heart and kidneys, for example, said Carr.

"If you have an agent that genuinely stays within the intravascular space, then it allows you to get a much more accurate measurement of perfusion. Extracellular agents, which leak out of vessels, may contaminate perfusion measurements," he said.

While blood pool agents' practical merits are still being evaluated, Meaney is optimistic.

"If they are just as good as first-pass agents for looking at any territory, then I would choose to use them because you can get a second bite of the cherry," he said. "That would be useful everywhere."