PCa Commentary
 

These three techniques exploit significantly different biologic aspects of tumor and host. The image in the bone scan results from the detection of the radiotracer on the mineralizing surface of bone, rather than in the tumor itself, and reports the host's reactivity, i.e, bone remodeling. Utilizing a parallel mechanism but by employing a different tracer, the NaF-18 PET detects the deposition of fluoride at sites of high bone turnover and remodeling where F-18 exchanges with hydroxyl groups in hydroxyapatite crystal of bone, to form fluoroapatite (Seminars in Nuclear Medicine,Jan 2001). MR imaging detects information about the tumor cells themselves by sensing chemical characteristics of cells, especially differences in water content.

One study in particular was instructive on a specific issue, "Sensitivity in Detecting Osseous Lesions Depends on Anatomic Localization: Planar Bone Scintigaphy Versue 18F PET", Schirrmeister, J Nuc Med, Oct 1999. This study, as in many others, presents a comparison between techniques by reporting detectability on a lesion-to-lesion basis. Because bone metastases in prostate cancer usually are osteoblastic/sclerotic as opposed to osteolytic, it's advisable to restrict comparisons of techniques to studies imaging prostate cancer (as opposed to thyroid or lung cancer), even though it is recognized that there is a significant osteolytic component to PC bone metastases. In the prostate cancer patients, NaF-18 PET detected 67 lesions and bone scanning 33 lesions. Again, using a composite of MRI, CT, and X-Rays as the reference "gold standard", on a lesion-to lesion basis the detection rates of bone metastases were 100% with NaF-18 in patients with osteoblastic metastases associated with cancer of the prostate. This was in contrast to detection with bone scanning where only 49% of the osteoblastic lesions were detected. When considering specific regions of the skeleton,... "Compared with PET and the reference methods, [bone scans] had a sensitivity of 82.8% in detecting malignant and benign lesions in the skull, thorax, and extremities and a sensitivity of 40% in the spine and pelvis." The costs for a Tc-99m bone scan is about $250 and a NaF-18 PET and Tc-99m SPECT bone scan are roughly the same at about $750. The medicare reimbursement for an MRI "marrow study" covering skull to pelvis plus ribs and femurs is $3600, and for a non-contrast MRI of thoracic and lumbar spine and pelvis, about $1800. The many studies of whole body MRI scanning have shown that the MRI is more sensitive than planar bone scans for detection of metastases. When the comparison between MRI and planar bone scans is restricted to axial spine and pelvis, the MRI is more sensitive, although bone scanning augmented by SPECT of these areas has comparable sensitivity to MRI.

Where does this leave us with respect to the initial evaluation of the "high risk" prostate cancer patient - the focus of this article? The greater sensitivity of MRI and NaF-18 PET for the detection of bone metastases seems well established. How does this sensitivity information apply in relation to the PSA value of a man at initial diagnosis?

Many studies argue against performing bone scans on men with initial PSAs of < 10 ng/mL, where the positive rate may be as low as 1-2%, with the possible exception of cases with extensive high grade tumor. A study, "MRI of the Skeleton in Prostate Cancer Staging", Scandanavian Journal of Urology and Nephrology, 37(3), 2003, presents information about 76 men correlating the initial PSA level with the results of MR imaging of the lower thoracic and lumber spine and pelvis. In the cohort with PSA < 20 ng/mL 4/27 (17%) were positive; and in the group with PSA > 20 ng/mL 22/52 (42%) were positive. The same authors in 1999 had done a previous study of 446 men in which they related the initial PSA with bone scan positivity. Results: PSA <5, 4.5% positive; PSA 5-9, 5.2%; PSA 10-19, 3.2% and 20-49, 37%, PSA 50-99, 50.9%. When they factored in tumor grade, bone scan positivity was much less in grade 1 and 2 tumors. The authors then compared the data from both studies. They averaged the results from from the bone scan study for PSA <5 to 19: = 4.3%, which was then compared to the 17% pickup in the men in the same PSA range in the MR study. Their conclusion: "...MRI is a more sensitive indicator of bone metastases than bone scintigraphy in the low range of serum PSA, but less sensitive in the high range." [However, the latter portion of the conclusion is weakened because the MR imaging covered only the axial spine and pelvis whereas bone scanning examines the entire skeleton]

Its important to recognize that the pathway of hematogenous (venous) spread from prostate to bone under situations of increased intraabdominal pressure (coughing, straining) can bypass the caval tributaries and flow preferentially into the pelvic bones and the entire spinal axis via Batson's vertebral venous system.Perhaps this explains why isolated asymptomatic metastases in the peripheral skeleton are very rare (Traill, Clin Radiol 1999;54:448-51). In their comparative study of axial MR versus bone scanning in 200 patients with breast and prostate cancer only 4/200 (2%) had isolated lesions in the peripheral skeleton and 3 of these 4 were symptomatic. Seattle Nuclear Medicine is considering offering the NaF-18 PET in Spring of 2004, but insurance coverage is not yet available for this study in prostate cancer. After considering the sensitivity data presented, and until whole body MRI or NaF-18 PET become available, a clinician might well choose a spinal axis and pelvic MRI, as opposed to planar bone scanning in the initial work-up of the high risk patient.

The formulation of this article is solely mine, but I deeply appreciate the constructive comments and suggestions for references by Drs. Justin Smith, Udo Schmiedl, and Dave Haseley - Ed Weber

Bottom Line: Advances in imaging techniques can enhance staging accuracy for prostate cancer.

« Back to Article List