In prostate MR spectroscopy (MRS), the citrate-to-choline concentration is a magic ratio for detecting metabolic changes in cancer. Specifically, the concentration decreases in cancerous prostate tissue. But on MRS, the citrate signal suffers from poor resolution, mostly because of a strong coupling with choline, according to presenters at the 2006 International Society for Magnetic Resonance in Medicine (ISMRM) in Seattle.
The researchers, from the Institute for Biomedical Engineering at the University of Zurich in Switzerland, performed prostate spectroscopy at 3-tesla using 2D S-PRESS, and found they were able to establish an optimal echo time while yielding the entire spectral information in a single measurement.
"Two dimensional J-resolved spectroscopy (JPRESS) is capable of resolving the coupling information and has thus gained popularity for prostate exams," wrote Thomas Lange, Ph.D., and colleagues in their ISMRM poster presentation. Lange was one of the co-authors of a recent paper on improved 2D J-resolved spectroscopy (NMR in Biomedicine, April 2006, Vol. 19:2, pp. 264-270).
"Another approach for the detection of strongly coupled spin systems is the difference-editing technique S-PRESS," they added. In their poster, they introduced a 2D technique for the detection of strongly coupled metabolites called 2D S-PRESS.
They performed their study in phantom experiments and in 11 healthy human prostates on a 3-tesla whole-body scanner (Achieva, Philips Medical Systems, Andover, MA). The solution for the phantom experiment consisted of 15 mM choline (uncoupled), 30 mM lactate (weakly coupled), and 50 mM citrate (strongly coupled). For JPRESS, the TE range was 41.58 and 438 msec. For S-PRESS, the TE was 438 msec. In the human subjects, images were acquired using a two-element surface coil and the TE was 282 msec.
According to the results of the phantom studies, "the lactate resonance (weakly coupled) only shows a splitting in the indirect dimension for JPRESS, but not for 2D S-PRESS," the authors wrote. "Citrate (strongly coupled) gives rise to a relatively complicated peak pattern (in) JPRESS, while in the 2D S-PRESS spectrum there are two characteristic 'strong coupling' doublets.... The 2D S-PRESS technique enables an improved resolution and characterization of the citrate signal."
The authors concluded that their approach was superior to a one-dimensional difference-editing technique. Also, their study results suggested that 275 msec was the optimal echo time for 2D S-PRESS in prostate spectroscopy.
Transition zone cancers
In a second MR prostate study, a group from Memorial Sloan Kettering Cancer Center in New York City used the modality to detect, localize, and stage transition zone prostate cancers. On a digital rectal exam, transition zone cancers are more difficult to find than peripheral zone cancers, according to Dr. Oguz Akin and colleagues from the departments of radiology, epidemiology and biostatistics, and urology.
While only 25% of prostate cancers occur in the transition zone, these patients do have a more favorable prognosis. "There is a growing demand for further individualization of treatment plans, which necessitates the accurate characterization of the location and extent of cancer," the group wrote (Radiology, June 2006, Vol. 239:3, pp. 784-792).
This retrospective study consisted of 986 patients with at least one tumor in the transition zone. All patients underwent MRI and radical prostatectomy. MR was performed on a 1.5-tesla scanner (Signa, GE Healthcare, Chalfont St. Giles, U.K.) with a body coil for excitation and a pelvic phased-array coil, as well as an expandable endorectal coil (Medrad, Indianola, PA). The acquisition protocol included transverse, spin-echo T1-weighted imaging and thin-section, high-spatial resolution transverse, coronal, and sagittal T2-weighted fast spin-echo images.
Akin and co-author Dr. Evis Sala, Ph.D. -- both of whom were fellows at the time of the study -- interpreted the images under the guidance of co-investigator Dr. Hedvig Hricak, Ph.D. Homogenous low T2 signal intensity, ill-defined margins, and lack of capsule were considered indicative of transition zone prostate cancer on MR. The readers used a five-point scale to judge the location of the cancer.
According to the results, there were 223 transitional zone cancer foci, with a median volume of 0.77 mL, in 148 patients. Eleven percent of these cancers had extraprostatic extension.
For detecting the presence of transition zone cancers, reader 1 turned in a sensitivity of 75% and a specificity of 80%. Reader 2 posted a sensitivity of 80% and a specificity of 78%. Reader 1 correctly identified 56% of transition zone cancer foci with six false-positive findings. Meanwhile, reader 2 found 63% with 10 false positives.
For detecting extraprostatic extension, reader 1 had a sensitivity of 56% and reader 2 had a sensitivity of 28%. The specificity was 94% for reader 1 and 93% for reader 2.
The area under the curve (AUC) for reader 1 was 0.75 and 0.73 for reader 2. Inter-reader agreement was fair, the authors stated. For detecting cancers with a volume of less than 0.77 mL, reader 1 had an AUC of 0.67 and reader 2 had an AUC of 0.66. For cancers over 0.77 mL, the AUC was 0.84 for reader 1 and 0.82 for reader 2.
"Our findings show that the accuracy of transition zone cancer detection at MR imaging is related to the transition zone cancer volume, with higher accuracy for cancers with larger volumes," the authors wrote.
In addition to tumor volume, and the characteristics used in this study to delineate cancer, they also found two other indicative traits: lenticular shape and invasion of the anterior fibromuscular stroma. Taken altogether, "it is possible to detection transition zone cancers at MR imaging.... These MR imaging features need to be validated with an independent data set by a different group of reviewers," they stated.
Finally, another MR prostate study is slated for the September issue of the American Journal of Roentgenology. This article will focus on a combined protocol for MRI and MRS, resulting in a shorter exam time. It will also look at the predictive value of the two modalities for prostate tumor extent, extracapsular extension, and staging performance.
By Shalmali Pal
AuntMinnie.com staff writer
June 14, 2006
Related Reading
Low-grade prostate cancer may not require aggressive treatment, May 26, 2006
Prostate cancer screening not supported for older men, May 2, 2006
3-tesla MR shows promise for abdominal, pelvic imaging, April 4, 2006
Low-field MR with the right coil tops 3-tesla MR for prostate cancer imaging, December 15, 2005
Copyright © 2006 AuntMinnie.com
