Fusion imaging gains ground in RFA

Fusion imaging is making inroads as an imaging strategy for RFA and other tumor ablation techniques. The sophisticated melding of functional and structural imaging modalities allows physicians to more accurately map out the procedure beforehand and monitor the results afterward. A few software packages allow physicians to check their real-time progress against preprocedural images, and new products and procedures to automate the process are being developed.

CURRENT PRACTICE

Fusion imaging is necessary because no one imaging modality is perfect. CT can't always identify what tissue has been treated during a procedure, and it may not outline a tumor without contrast. Ultrasound falls prey to microbubbles during a procedure; the image can become obscured by steam or vapor from the heated tissue. Physicians can get around these problems by comparing more detailed preprocedural images with the image they see during the procedure.

Preprocedural PET/CT scans help physicians plan their strategy. The combination of CT, which demonstrates the tumor's structure, and PET, which shows its metabolic activity, provides a clearer view of just what the physician is facing.

Dr. Jason Williams of the American Cancer Ablation Center in Gulf Shores, AL, prefers to use PET/CT to map out a tumor site before a procedure to look for additional tumors that may have escaped notice. In addition, PET/CT can identify tumors that have been treated with radiation, leaving only a portion of the larger mass still active.

"It's as if you have a pizza," said Dr. Angel Williamson, who performs PET/CT imaging for Williams' patients at her Pensacola, FL, clinic, Angel Williamson Imaging Center. "If you do CT, it will call the entire pizza cancer, but if you do PET/CT, you will find out only the pepperoni is cancerous. You don't need to ablate the whole thing."

PET/CT can also be useful for postprocedure monitoring. Physicians using CT to check up on an ablated tumor must wait to see if the treated tumor grows before they can call the treatment successful, a process that may take up to a year. Using PET/CT, Williams can check the site for metabolic activity within a few months and re-treat if necessary.

PET/CT isn't perfect, however. The modality may be prone to false positives, as PET can read any area of inflammation or infection as "hot," or malignant, Williamson said. She makes it a practice to go over the scan with the patient to see whether there are injuries that can explain any suspicious findings.

In addition, a PET/CT scanner is so expensive that even some proponents don't feel the need to acquire their own. Williams, for example, prefers to send his patients to other providers for their scan rather than buy his own PET/CT scanner.

WHAT'S NEXT

A few physicians and equipment manufacturers have taken fusion imaging a step further, using preprocedural CT or MRI data sets simultaneously with real-time ultrasound imaging of the tumor site.

"When you're doing the ultrasound scan, you see a screen where, as you move the probe and look at oblique or off angles of the liver, for example, you'll see the corresponding CT image moving at the same time," said Dr. Matthew Callstrom of the Mayo Clinic.

Some fusion imaging tools enable physicians to combine PET with CT or MR images. Using Siemens Molecular Imaging's Miraview Target product

(formerly CTI Mirada Solutions), for instance, users can load any two modalities, fuse them, adjust them, define treatment contours, label the contours, and transmit them to a treatment planning system, said Dr. Joel Leong, executive medical director.

The next generation of fusion imaging tools for tumor ablation is in development. Dr. Brad Wood of the National Institutes of Health is working on an "interventional suite of the future" that combines ultrasound, rotational angiography, multislice CT, therapeutic ultrasound, and robotic automated or semiautomated needle placement or guidance.

The first fruits of this research are already being seen. An article in the April 2005 Journal of Vascular and Interventional Radiology described a feasibility study on the use of electromagnetic tracking devices for navigation during interventional radiology procedures. The tip of the needle can be tracked inside the body and compared with an image from a preprocedural data set.

"It's like having a mini-GPS inside the needle," Wood said. "This tells you where the needle is and it moves with breathing. It's better than just tracking the ultrasound itself."

The next step is to register the intraprocedural ultrasound to a preprocedural CT, he said. An ultrasound scanner that can be used in a normal fashion is encoded and registered to a CT scanner, allowing the physician to switch back and forth between the modalities, taking advantage of each one's strengths.

Finally, Wood envisions a robot encoded to the CT scanner that knows where it is on the CT image and where the patient's tumor is. This would allow users to layer data: A PET/CT or MR image could be fused to the CT scan, which physicians could use as a point-and-click, see-and-treat process that is seamless, iterative, and provides feedback.

"Right now, we're evaluating how automated we want to get," Wood said. "A robot can pick angles better than any human can based on the preprocedural CT. And if you layer the data sets on top of that, you have a pretty powerful tool."