Whole-body imaging brings new slant to cancer staging

Radiologist interest in whole-body diffusion-weighted MRI for cancer applications is intensifying following preliminary studies that demonstrate its potential value for staging cancer. Whole-body DWI produces a composite image using a STIR echo-planar diffusion-weighted technique with a high b-value for background suppression.

During DWI, composite whole-body images are acquired at multiple stages before reconstruction. Maximum intensity projections are then often displayed in a reversed black-to-white gray-scale. Highly cellular lymph nodes showing restricted diffusion are especially conspicuous. Whole-body DWI adds only one to five minutes to the imaging time, and, unlike PET/CT, it involves no ionizing radiation.

Studies presented at the ISMRM/ESMRMB meeting found that whole-body DWI may outperform FDG-PET and is significantly more accurate than purely anatomic whole-body MRI for M-staging of lung cancer. Speaking for his colleagues at Kobe University Graduate School of Medicine in Japan, Dr. Yoshiharu Ohno presented the first prospective clinical trial comparing the accuracy of the three approaches to lung cancer staging. Ohno is director of functional and diagnostic imaging research at Kobe.

Subjects were 74 consecutive patients with confirmed lung cancer: 47 cases of adenocarcinoma, 13 of squamous cell carcinoma, and seven each of small- and large-cell cancer. With per-patient sensitivity and specificity rates of 85.7% and 90%, respectively, whole-body DWI was more specific than FDG-PET, which had sensitivity and specificity rates of 85.7% and 81.6%, respectively. The sensitivity and specificity of whole-body MRI without DWI were 71.4% and 85%, respectively.

Also speaking in Berlin, Dr. Chun Yan Zhang reported promising results in a preliminary study of 3T whole-body DWI performed on 152 consecutive patients with suspected prostate cancer at Peking University First Hospital in Beijing. Of 26 high-intensity bone lesions identified with DWI in six patients, 25 were subsequently confirmed to be bone metastases on conventional MRI. DWI also found 27 enlarged lymph nodes in four patients.

Dr. Michael A. Jacobs and colleagues at Johns Hopkins University School of Medicine demonstrated the feasibility of whole-body DWI with apparent diffusion coefficient mapping to identify and potentially characterize bony and soft-tissue masses.

Dr. Matthias Philipp Lichy and colleagues at the University of Tubingen in Germany employed a high b-value (1000 sec/mm2) single-shot echo-planar imaging approach to compare whole-body DWI with ADC mapping and PET/CT imaging.

Twenty cancer patients were imaged with whole-body DWI with ADC mapping on a 1.5T scanner. F-18 FDG was administered to 11 melanoma patients, four lung cancer patients, and one non-Hodgkin's lymphoma patient. Four prostate cancer patients were examined with carbon-11 choline.

Based on the PET findings, whole-body DWI detected 100% of the melanoma and lymphoma metastases. It found all the prostate lymph node metastases and lung metastases but was only 13% accurate for identifying lymph nodes containing metastatic cells from lung cancer.

Lichy still found room for technical improvement. Artifacts from spatial distortions were apparent with EPI in areas of high susceptibility, such as the bowel, and respiratory artifacts appeared in regions of interest in the thorax.