Contrast MRI Best at Finding Liver Trouble - But Timing Matters

For spotting the tiniest metastases inside the liver, MRI with the liver-specific gadolinium-based contrast agent Primovist (gadoxetic acid) was far and away the top choice of surgeons in an eight-country multicenter study.

The surgeons were given a chance to use various imaging methods for colorectal center patients and judge their value. The same contrast agent is also proving adept at visualizing obstructions in biliary vasculature, compared to non-contrast-enhanced MRI.

The 35-center VALUE study is well named, because its main goal was not to assess the sensitivity and specificity of Primovist-enhanced scanning in detecting liver metastases. Instead, the object of the study is to learn which of three imaging methods has the greatest impact on clinical decision making, at the best price, said Christoph Zech of the University of Munich, at ECR 2011 in Vienna, Austria, this week. The ultimate goal is to find a way to reduce the number of tests required for clinical decisions in this setting, and to eliminate the tough calls when surgeons come across unsuspected metastases and need to modify their plans ad hoc.

Based in Vienna, Berlin, and Bangkok, VALUE has now closed accrual at more than 350 colorectal cancer patients, but the interim results reported at ECR 2011 included analysis of the first 284. Even so, the results are definitive.

At the outset, patients were randomized to one of three imaging tests: multidetector CT, MRI with extracellular contrast, and MRI with Primovist. At post-scan consensus meetings, surgeons determined whether each patient required further testing. Ultimately the results will be compared in terms of surgical and patient outcomes, and cost. For the moment, we know what the surgeons saw.

In no case did a surgical team request further information after seeing a Primovist-enhanced MRI (PV-MRI) scans. But they did seek a second scan after 16 percent of extracellular contrast MRI images and after more than a third (36 percent) multidetector CT scans. In all but one case, surgeons chose PV-MRI for the second-look scan.

Furthermore, when surgeons were asked to judge their level of confidence in each scanning method, 98 percent gave PV-MRI a high or very high rating, compared with only 67 percent for multidetector CT. In one case, PV-MRI even detected a small liver metastasis that did not appear on intraoperative ultrasound.

"These results clearly show the superiority of PV-MRI," said Zech. Which scan would reduce utilization is clear; economic outcomes are forthcoming.

Doctors at the University of Pisa have also found Primovist useful in visualizing ischemia and stones in liver transplant patients. In a sample of 28 patients, said Piero Boraschi, Primovist-enhanced MRI "significantly tended to out-perform conventional MR cholangiography" at showing strictures in the hepatic bifurcation and the extrahepatic bile duct, as well as sludge and stones.
Images taken at five-minute intervals between 15 and 40 minutes after contrast injection were analyzed by consensus reading of two observers. The enhancement times appeared to correlate with the degree of biliary obstruction, which leads the team to propose that contrast-enhanced MRI of the liver may eventually be useful to provide not only anatomical, but functional information about the liver.

The method of timing contrast injection is crucial for accurate MRI studies of the liver, cautioned Christine Schmid-Tannwald, of the University of Munich. She began her remarks by showing three quite distinct MRI scans, depicting hepatocellular carcinoma (HCC), hemangioma, and focal nodular hyperplasia (FNH) of the liver. In each scan, the abnormality was visible in the same quadrant, but they varied markedly in appearance and resolution.

"It's clear that we need robust contrast not only for detection, but for characterization of liver malignancies," she said. What wasn't clear before the study was whether a fixed delay is sufficient to make the distinction, or whether it is necessary to carry out bolus timing.

The Munich team retrospectively evaluated scans from 46 patients referred for MRI-based detection of suspected liver cancer. These had been assessed in 3 different ways: 20 with the CareBolus detection method, 18 with a test bolus technique with transit time determined by "special software," and the third with a 20-second fixed delay. Two blinded interpreters separately scored the quality of the resulting scans according to the degree of enhancement in the aorta, the arteries, the portal and hepatic veins, and the kidney.

A combined rating score showed that the Care Bolus technique and the test software were about 50 percent more accurate than the fixed delay at visualizing internal structures in the liver. (Both scored an arithmetic mean of nine in the interpreters' scores; fixed delay scored only 6.2) In contrast-enhanced liver scans, concluded Schmid-Tannwald, bolus timing is "mandatory."