Report from ECR: Radiologists find role for whole-body MRI in spotting metastases

March 21, 2007

If patients suffering from malignant disease are to get the right treatment and an accurate prognosis, accurate assessment of metastases is crucial. Whole-body MR is a good tool that can play a supporting role for detection of metastases, but it is not as reliable as gold standard PET/CT, according to research presented at the European Congress of Radiology.

If patients suffering from malignant disease are to get the right treatment and an accurate prognosis, accurate assessment of metastases is crucial. Whole-body MR is a good tool that can play a supporting role for detection of metastases, but it is not as reliable as gold standard PET/CT, according to research presented at the European Congress of Radiology.

In one preliminary study of 12 patients, researchers at the University Clinic Essen put a new continuous table movement technique to the test with multicontrast whole-body MR. Typically, images are generated in sections and then combined using data processing techniques. With the new table technique, the patient is moved through the magnet, which has benefits for speed and quality.

"Head-to-toe imaging was done with whole-body MRI in less than 10 minutes, providing sufficient spatial resolution," said Dr. Florian Vogt, who presented study results.

Patients involved in the study had a history of a range of cancers, including colon, bronchial, and gastrointestinal tumors, and had been referred for PET/CT. Researchers obtained T1- and T2-weighted MR images and then performed contrast-enhanced MR scans. Findings were compared with PET/CT.

PET/CT identified 75 lesions. Of these, the whole-body MR sequences picked up 73.

"MRI detected all pulmonary, liver, and skeletal metastases. However, it missed two lymph node metastases picked up by PET/CT. MRI also detected an additional seven spine lesions that were missed on PET/CT," Vogt said.

Almost all MR-detected lesions were visible on the contrast-enhanced T1-weighted real-time scans, but some additional lesions were seen only on T2-weighted images. The research suggests the potential value of viewing both T1- and T2-weighted images, plus contrast-enhanced images, in boosting cancer detection rates.

In another study presented during the same session, Italian researchers concluded that whole-body MR is effective and fast for examining colon cancer patients, but it is still not as good as whole-body PET/CT.

Researchers at the University Tor Vergata in Rome recruited 20 patients who had undergone surgery for primary colon cancer and were referred for PET/CT for assessment of metastases between October 2005 and March 2006.

The researchers performed three different whole-body MR sequences on a 3T unit, as well as the indicated FDG-PET/CT study. It took about 23 minutes to perform whole-body PET/CT compared with nine minutes for whole-body MRI. Following imaging, patients were followed for an average of 186 days. Histological results were compared with imaging findings.

Both imaging techniques succeeded in finding lymph node involvement in 15 of 20 cases. In some cases, such as peritoneal implants, it proved difficult to image metastases with MR but possible with PET/CT.

When it came to distant metastases, overstaging was problematic with whole-body PET/CT, while whole-body MR had the opposite drawback of understaging some patients.

Overall, performance was better with whole-body PET/CT, according to Dr. Giovanni Squillaci, who presented results.

In a third study, researchers at the Eberhard-Karls-University in Tuebingen, Germany, performed a prospective blinded trial comparing contrast-enhanced whole-body MR with whole-body PET/CT. In 64 patients with advanced melanoma, there were 420 extracerebral lesions suspicious for malignancy.

PET/CT scored an overall accuracy of 86.7% for assessing lesions, outperforming whole-body MR (78.8%), CT alone (75%), and PET alone (74.3%).

MRI was less sensitive but more specific than PET/CT for lung metastases. All 37 liver lesions were correctly assessed by whole-body MRI, whereas PET/CT missed two liver metastases. Whole-body MRI proved superior in detecting bone metastases, with 100% sensitivity and 73.3% specificity, compared with 91.4% sensitivity and 80% specificity for PET/CT. Both techniques underestimated lung lesions smaller than 10 mm.

The results suggest that whole-body PET/CT is best suited as a first-line staging modality, but MRI has strong potential in a complementary role for assessment of liver, bone, and cerebral metastases.

For more online information, visit Diagnostic Imaging's ECR 2007 Webcast.