SPECT is far from the new kid on the molecular imaging block, but except for cardiac imaging, this nuclear medicine modality has not yet realized routine clinical utility as a replacement for conventional planar scanning. Recently developed reconstruction protocols and hybrid SPECT/CT scanners, however, could breathe new life into SPECT by opening the door to clinically feasible whole-body SPECT/CT. The University of Texas M.D. Anderson Cancer Center in Houston, is at the cutting edge of SPECT/CT research. This month, M.D. Anderson experts, Chair of Nuclear Medicine , Associate Professor , and Senior Medical Physicist Bill Erwin from the department of Imaging Physics discuss the clinical path to and potential of whole-body SPECT/CT.

The University of Texas M.D. Anderson Cancer Center Department of Imaging Physics
Homer Macapinlac, MD Chair of Nuclear Medicine	Osama Mawlawi, PhD Associate Professor	Bill Erwin Senior Medical Physicist

Has nuclear medicine optimized SPECT/CT?

Macapinlac: No, definitely not. If one uses PET/CT as a model, one can see the potential of SPECT/ CT. The combination of CT and PET data produces a more complete image set, providing the physician with the anatomical and functional views necessary to render a solid diagnosis. The goal of SPECT/CT is to utilize a system capable of multislice CT scanning to improve the interpretation of SPECT studies. Standard SPECT studies can lack specificity; the addition of CT seems to add data needed to boost specificity.

What factors limit the utility of SPECT/CT?

Macapinlac: Two challenges confront SPECT/CT and limit its value in routine clinical use. For starters, the Centers for Medicare and Medicaid Services (CMS) has not approved reimbursement for the CT part of the study, which presents an economic challenge. On top of that, the time to acquire a SPECT study reduces its clinical value. It can take from 20 to 45 minutes to acquire a single-bed SPECT study, making multi-bed scans not tolerable for many patients. Consequently, SPECT/CT studies are typically limited to a single anatomical area such as the chest. In some cases, a second or third study of the abdomen or pelvis is performed later.

How does whole-body SPECT/CT differ from localized studies?

Mawlawi: Whole-body SPECT/CT requires three to five bed positions to cover the entire body. Traditional step-and-shoot acquisition completes an image at one angle and then pauses acquisition counting before moving to the next position to acquire data. As a result, it might take anywhere from 60 minutes to two hours to acquire a whole-body SPECT/CT.

Can you describe the whole-body SPECT/CT research program at M.D. Anderson?

Macapinlac: We’d like to determine if we can complete a whole-body SPECT/CT study using multiple bed positions in a clinical environment. Our physics team is trying to determine if it can link together multiple bed studies to reconstruct the scans and transform SPECT/CT into a routine acquisition protocol.

Our goal is simple; we want to mimic PET/CT with SPECT/CT. That is, we’d like to acquire the cross-sectional images in a single pass to make the entire whole body study tolerable for patients, which requires an acquisition time in the 25 to 30 minute range.

Mawlawi: Our vision is to transfer planar bone imaging to SPECT/CT. Currently, 40 percent of our patient load is bone imaging, which primarily consists of planar imaging. SPECT/CT can deliver several advantages including improved sensitivity and specificity. The hitch is completing the study in a time slot comparable to planar images, or about 25 minutes.

The 25-minute goal is attainable with a six to eight minute acquisition time per bed position. Our team researched the potential by acquiring SPECT scans of different durations including four, six, eight and 10 minutes and asking multiple physicians to compare the image quality of the scans. We imaged 40 patients on Siemens Medical Solutions e.Cam SPECT and TruePoint Symbia SPECT/CT systems. The consensus seems to be that seven minutes per bed position will produce a scan with optimal image quality in the appropriate length of time.

Are other advances facilitating this approach?

Macapinlac: New SPECT systems reduce the acquisition time, which makes it feasible to acquire the CT study. In addition, our physics research team is working with Siemens to develop software to help improve the reconstruction process. Oncoflash 3D ordered-subset expectation-maximization (OSEM) iterative reconstruction software is an important development. First, reconstruction times are greatly reduced with Oncoflash [approximately four times faster than current options], allowing an entire whole-body SPECT scan to be reconstructed both with and without attenuation correction, as is done for PET,  in under 10 minutes. Second, Oncoflash improves the reconstruction of slices along the top and bottom of the field-of-view for each bed position, which are known to be problematic for 3D reconstruction algorithms. This will improve the quality of the whole-body SPECT dataset when it is formed by combining the individual bed position reconstructions [so-called “stitching” operation].

Mawlawi: Most new scanners can operate in the conventional step-and-shoot mode or in continuous acquisition mode. To complete the scan in a reasonable amount of time, the camera must operate in continuous acquisition mode, to eliminate the dead time between views that would occur in step-and-shoot mode and retain all counts, which is vital to maintain image quality when reducing acquisition time.

Does the change in acquisition mode require a different approach to reconstruction?

Erwin: We looked at different iterations and subsets. Our reconstructions are full 3D iterative and take into account attenuation, scatter and system resolution. In this case, we opted for a reconstruction with eight iterations and 15 subsets with 8 mm Gaussian post-filtering. The reconstruction parameters can be further optimized, once an acquisition time is settled upon.

Can you describe how this might impact patient care?

Macapinlac: The potential is very intriguing. Whole-body SPECT/CT could improve the way we interpret bone scans. The current protocol is intense and time-consuming. If we see a suspicious finding, we order a SPECT scan and plain film. If those studies are negative, the next step is a CT or MRI. Whole-body SPECT/CT streamlines the protocol, limiting the studies to a plain film and SPECT/CT. With SPECT/CT, the physician can determine if a finding is benign or malignant in a single pass. In addition, it appears that the correlative images provide better specificity than conventional SPECT. (Large clinical trials demonstrating the improved specificity of SPECT/CT aren’t yet available.) The approach could better inform physicians and provide them with data needed to refine treatment for individual patients. For example, if a patient is scheduled for a mastectomy, the whole-body SPECT/CT could inform the surgeon and oncology team if the disease metastasized to the bone prior to surgery.  

Mawlawi: Ultimately, whole-body SPECT/CT offers a means to increase physician confidence. They will have a clearer understanding of the extent of disease.

It sounds like whole-body SPECT/CT could broaden the role of SPECT in oncology.

Macapinlac: Absolutely. The promise of SPECT/CT exists not only on the diagnostic side, but also on the therapeutic side. Whole-body SPECT/CT could be used to quantify the distribution of an internal radionuclide therapy treatment dose, which represents a significant improvement over the conventional methods that rely on planar images and “guesstimates.”

Where do you see SPECT/CT headed in the next few years?

Macapinlac: Right now, whole-body SPECT/CT is not quite ready for prime time. I believe we will see multiple factors converge over the next few years to bring these scanners into routine clinical use. Medicare approval, well-designed reconstruction software and properly developed and researched protocols can move SPECT/CT into the clinical environment.