Pilot Project: Fast Whole-body Spect Scanning to Improve the Detection of Bone Metastases in Patients With Diagnosed Cancer
Bone scintigraphy (e.g. "bone scan") with Tc-99m diphosphonates is one of the most common
imaging radioisotope procedures in departments of nuclear medicine worldwide, with several
thousand performed every year in a busy nuclear medicine service. Despite giving fast and
accurate skeletal surveys in conditions ranging from metastatic cancer, to fractures, to
infection, the technology has seen little change since the early 1970's. The Anger camera
remains the instrument of choice for taking serial images of a patient's skeleton, and since
the late 1970's SPECT (Single Photon Emission Computed Tomography) - essentially "nuclear CT
imaging of 3-D distribution of radiotracer" has been added to improve contrast resolution
and lesion placement.
Despite the known greater sensitivity of SPECT compared to planar bone nuclear medicine
studies, SPECT is usually done only in limited patients due to time constraints. This is
because a standard tomographic bone acquisition requires about 20 minutes, therefore it
usually can only be performed over one body region such as the abdomen or chest.
Skeletal PET scanning with F-18 Fluoride has become popular since the introduction of
multiple PET scanners for imaging cancer with F-18 FDG. It provides much higher count rates
than traditional bone scans done with Tc-99m phosphates, and many authors have shown that it
is the most sensitive imaging modality for detecting skeletal lesions. Indeed, several PET
centers have proposed replacing the standard Tc-99m bone scan with F-18 PET scans. However
the short half-life of F-18 makes this type of study practical only for PET centers within a
short distance from a cyclotron. Furthermore, most PET scanners are usually fully occupied
doing other types of diagnostic oncology studies, and the cost of the F-18 is currently not
reimbursed for simple skeletal surveys.
Recently it has become feasible to dramatically shorten the time necessary for acquisition
of a nuclear medicine SPECT study. This has occurred with the introduction of new image
reconstruction software which includes resolution recovery correction (so called "collimator
de-blurring"). Since this approach results in significant improvement of image resolution,
many centers are currently looking at the feasibility of using this new technique to reduce
imaging times. For example, its use has been investigated for nuclear cardiac scans with
acquisition times shortened by half.
We propose to investigate the performance of the image reconstruction software with
resolution recovery correction for bone SPECT studies. We estimate that in only 30 minutes,
using this new technique of collimator de-blurring, one could perform a fully 3-dimensional
SPECT whole-body bone study, essentially obviating the necessity for doing planar bone
In the scope of the proposed project, our group aims to test the hypothesis that one can
perform a Tc-99m whole-body SPECT study in the same time as a regular routine planar bone
study, with greater localization accuracy, and greater lesion detection.
To establish a "gold standard" necessary to assess the performance of the SPECT bone scans,
we will compare number of malignant lesions detected in patients who are proven to have
metastatic skeletal bone lesions on PET F-18 whole-body scans, with whole-body Tc-99m SPECT
lesions. We also propose to compare the detection of SPECT scans with standard planar bone
scans. This will allow for two major comparisons (a) the accuracy of SPECT bone studies
compared to planar bone studies, and (b) the accuracy of SPECT bone scans compared to F-18
PET studies. Most prior studies purporting to show the superiority of F-18 bone scans to
Tc-99 bone scans were done only against either planar scans or a combination of planar scans
and partial SPECT studies over the spine. We anticipate that F-18 bone scans, due to the
higher counting statistics of PET agents, will show more lesions than SPECT, but the exact
increase in sensitivity has never been compared to whole-body SPECT scans.
1. Evaluate the feasibility of implementing whole-body SPECT in routine practice with
standard SPECT gamma cameras
2. Evaluate the value of whole-body SPECT compared to whole-body planar studies for the
detection of bone metastases
3. Compare the results of state-of-the art whole-body SPECT skeleton surveys with
state-of-the art whole-body Na18F PET whole-body bone surveys.
4. Obtain preliminary data on the respective accuracy of planar bone scintigraphy,
whole-body SPECT and Na18F PET/CT to plan future studies on the cost-effectiveness of
these methods to improve prostate cancer staging.
1. A whole-body bone SPECT is more sensitive and specific than a conventional planar bone
scan for the detection of bone metastases
2. Whole-body bone SPECT is less sensitive than a whole-body bone PET but the lower
sensitivity has minimal impact on the overall detection of bone metastases on a patient
Overall study design
This is an open phase II study designed to accumulate preliminary data on the relative
diagnostic value of planar bone scintigraphy, whole-body bone SPECT and whole-body bone PET
with sodium 18F-Fluoride. We plan to accrue a total of 25 subjects with suspected bone
metastases who will undergo planar bone scintigraphy, bone SPECT and a Na18F whole-body bone
All patients studies will be analyzed by 3 independent nuclear medicine readers, with number
of lesions identified on planar scans being compared to number of lesions on non-Astonish
SPECT bone scans and Astonish SPECT bone scans.
Each study will be scored by noting:
- number and localization of lesions detected
- level of certainty (1 = definitely malignant, 5= definitely benign)
- general visual quality of images (graded from 1 = poor to 5 = excellent)
The Na18F PET/CT bone study will serve as the reference gold-standard study against which
the other acquisitions will be judged.
This is a pilot study which will provide preliminary data for further power calculations to
estimate the study size needed to demonstrate a statistically significant improvement of
whole-body SPECT compared to planar scintigraphy. With 25 subjects, we will have 82% of
chance to demonstrate a statistically significant 30% difference in accuracy between planar
scintigraphy and whole-body SPECT compared to the reference standard (PET/CT).
Observational Model: Case-Only, Time Perspective: Prospective
Anna Celler, PhD
University of British Columbia
Canada: Health Canada