May 5 , 2008
Refined Viewing — In-Room Imaging Brings New Style to Radiation Therapy
By Pamela Kropf
Vol. 9 No. 8 P. 14
In-room and onboard imaging raise interesting questions about improving accuracy, dose, and quality control.
Diagnostic imaging modalities enable oncologists to obtain highly accurate images of tumors and their surrounding tissue at any point in time. But because tumors and surrounding physiology change, researchers and clinicians are developing ways to utilize imaging at the time of a radiation treatment session. Their goal is to more precisely target where to treat and confirm that the planned radiation dose is delivered where intended.
These “in-room” and “onboard” techniques help account for tumor changes, other physical changes, and/or patient movement. And these approaches potentially allow for a more accurate dose delivery than relying on one set of images throughout an entire treatment course.
Jean Pouliot, PhD, a professor in the University of California, San Francisco’s (UCSF) department of radiation oncology, which is part of the UCSF Comprehensive Cancer Center, says a regular CT scanner placed in the treatment room combines the benefit of proven imaging technology and the immediacy of imaging right before treatment. “Those [CT systems set up in the treatment room] are mature yet evolving devices with a good efficiency,” he says. “[But] on-board imaging systems are still in their infancy. Moreover, as they are attached to the gantry of the treatment unit and move around the patient, they are limited by regulation to one rotation per minute.”
Although in-room CT is faster and allows for more accurate dose delivery, the images themselves are not as accurate as planning CT images. It’s also more expensive and requires more quality assurance (QA), although there doesn’t seem to be a consensus about how much QA is enough.
Regarding tumor movement, Pouliot says most movements occur over days or minutes and for those, in-room CT is adequate. But for the subset of cases with fast movements such as respiratory motion, specific approaches are required.
Fang-Fang Yin, PhD, a professor and chief of medical physics in the department of radiation oncology at Duke University School of Medicine in Durham, N.C., says “four-dimensional [4D] CT considers organ movement but not real time. Everyone breathes differently at different times unless [the person] is extremely well coached. Typically, a 4D scan takes longer; it’s not optimal for the patient.”
There are now several studies showing that margins can be reduced in regard to more accurate dose delivery, according to Pouliot. “Also, with imaging, the confidence level of delivering the dose correctly is raised, allowing more aggressive treatments to be considered,” he adds. However, before images are taken, it’s important to note the purpose of acquiring the image.
“For patient alignment purposes, onboard CT images are adequate and accurate,” Pouliot says. “As real-time adaptive strategies are developed, automatic resegmentation of the organs and tissue deformation algorithms for nonrigid registration will play an important role. This is difficult to achieve with the current image quality of onboard CT, but the same can be said with planning CT images.”
But how much QA is enough? According to Pouliot, it should be interpreted in a more general sense. “If a patient recurred a few years after the treatment, wouldn’t we want to be able to correlate the recurrence site with the dose delivered to this patient so plans could be improved for other patients in the future?” he asks. “So QA is for validating that a treatment was correctly delivered and for properly archiving for future references.”
Yin notes that because in-room CT requires higher geometric accuracy relative to the treatment unit, more QA is needed.
Onboard imaging may be an extension of in-room CT, but it encompasses several modalities, including digital tomosynthesis (DTS) and megavoltage CT, also known as helical tomosynthesis. Pouliot says different imaging modalities are required to perform the different tasks related to accurate dose delivery.
“In-room CT [and/or onboard CT] can efficiently perform patient positioning and organ targeting. However, as dose distributions become more tightly conformed to the targets and margins are reduced, any change of anatomy due to weight loss, tumor shrinkage, or organ displacement due to breathing or organ filling require constant verification of the proper position with specially adapted imaging modalities,” Pouliot says. “The characteristics of DTS are intermediate between 2D radiography [portal imaging] and 3D CT. Compared to portal imaging, DTS improves tissue contrast by resolving overlay anatomy into slices; compared to onboard CT, it offers shorter acquisition time.”
Yin says onboard imaging will improve positioning accuracy by correcting systematic and random patient positioning errors. It also allows fluoroscopic imaging to monitor the internal target motion with a surrogate. However, he says onboard imaging time could affect treatment efficiency.
According to Pouliot, movement can imply different scales of time. He says displacement can occur over several days, minutes, or within seconds, so the technology must be adapted accordingly. “By reducing its acquisition time, an imaging system increases the range of clinical applications,” he says. “Movements on the scale of minutes occur for almost all patients, requiring fast imaging. Also, since the imaging time adds to the treatment time, shortening it is of real interest.”
QA encompasses the correct equipment performance, as well as the correct dose delivery to the appropriate organs, as planned, according to Pouliot, who says that many standard protocols and task group recommendations exist for the equipment.
“However,” he adds, “little is in place to document the dose delivered to the patient. There is a strong need to validate the actual delivered dose and to record a complete description of the 3D dose distribution received by each patient.”
While some facilities manage their own QA programs, others rely on outsourcing organizations that specialize in such tasks. For example, to help facilities meet state and Joint Commission QA standards, National Physics Consultants (NPC) provides comprehensive quality control programs to ensure such standards are met, the company’s Web site explains. NPC employs eight consultants who are experienced in diagnostic radiology and nuclear medicine to check for appropriate QA standards in the field. The NPC consultants assess quality in the following areas:
• diagnostic x-ray equipment performance, safety, output measurements, and tube performance criteria;
• radiation safety procedures;
• fetal dose calculation;
• x-ray room shielding design and evaluation;
• acceptance testing of radiographic and fluoroscopic units;
• personnel dosimetry reports to test the effectiveness of radiation safety practices;
• radiation safety policies for compliance with The Joint Commission; and
• procedure doses for radiographic units.
More specifically, NPC consultants look for CT QA in the following areas:
• beam alignment;
• slice width and scan increment accuracy;
• determination of contrast scale;
• degree of linearity of CT number response and system noise;
• high and low contrast spatial resolution and field uniformity;
• dosimetry measurements with respect to location in phantom, peak kilovoltage, milliampere, scan time, slice thickness, and scan diameter; and
• acceptance testing of CT equipment.
Tried and True Techniques
Yin says his facility uses onboard and in-room imaging techniques as needed, mainly based on accuracy requirements for treatment. So far, these techniques have improved accuracy for delivering advanced treatment, especially for hypofractionated stereotactic radiotherapy and intensity-modulated radiation therapy (IMRT), he adds.
The radiation oncology team at UCSF has developed a multiple adaptive planning strategy for IMRT for the simultaneous treatment of prostate and lymph nodes, Pouliot says. However, in this case, the problem is the independent movement of the prostate relative to the nodes.
“Onboard CT imaging is used daily to correctly determine the setup position based on bony anatomy and then the prostate position is determined from the same images by registering them on the implanted gold markers,” he says. “The shift of the prostate relative to the bony anatomy is computed to select one treatment plan among a series of precalculated plans reflecting the different possibilities of prostate displacement. This way the patient receives each day a treatment that is adapted to the location of the prostate while the nodes are adequately treated.”
Similarly, UCSF is working with DTS to exploit its specific characteristics. Pouliot says DTS has potential advantages in imaging lung patients in the treatment position because it has the ability to image the patient during a breath hold, minimizing the effect of tumor blurring due to respiratory motion.
In the Pipeline
Pouliot says that as 3D images of a patient in treatment position—seconds before the dose is delivered—become available, they demonstrate the variability of the anatomy.
“They also question the validity of basing the entire course of a treatment on a single CT image acquired days or weeks before. The availability of the 3D images allows also the recalculation of the 3D dose distribution” Pouliot states. “This is the enabler of DGRT [dose-guided radiation therapy], and although the applications and consequences of DGRT are only emerging, they are already numerous.”
So will radiation oncology eventually move from image-guided radiation therapy (IGRT) to DGRT? Yin is cautiously optimistic. “IGRT is the basis for DGRT, and they really try to achieve the same goal for radiation therapy,” he says. “A lot more work is needed to achieve DGRT.”
Pouliot says the two concepts are intertwined. “IGRT and DGRT are both adaptive strategies. IGRT ensures that the anatomy (patient position and organs) is positioned as it was intended. DGRT ensures that the anatomy is correct and that the dose delivered is in precise agreement with the dose plan,” he says. “Therefore, DGRT includes IGRT and is more stringent. DGRT uses the end point of radiation therapy [the delivered dose] instead of a surrogate [the patient position] to verify [and correct if necessary] the adequacy of the dose delivery process.”
“Furthermore,” he continues, “the ability to recalculate the dose distribution on the current anatomy [dose of the day] of the patient, as opposed to when the plan was prepared days or weeks before, will provide a level of quality assurance not yet available. It will become necessary to document that the dose plan was actually delivered to the patient. If only for this documentation purpose, DGRT is bound to become routine.”
— Pamela Kropf is a freelance writer based in King of Prussia, Pa.