LOS ANGELES—When positioning patients for external beam partial breast irradiation (PBI), body surface mapping is more accurate than either alignment using lasers or alignment using kilovoltage (kV) on-board imaging, a new study shows.
“As we transition from whole-breast irradiation to partial breast irradiation in selected patients, we worry about the increased risk of target miss and possible subsequent tumor recurrence,” Sasha H. Wahab, MD, said at ASTRO 2007 (abstract 43). “We wanted to look at various newer image guidance methods to see if we could identify one method that would allow us to improve the setup accuracy, compared to alignment with lasers alone.”
Dr. Wahab and his colleagues at Washington University School of Medicine compared three methods for aligning the treatment target for PBI:
• Conventional alignment using the treatment room lasers.
• 3D surface mapping using AlignRT (Vision RT).
• Matching of orthogonal kV images to skeletal anatomy, mainly the vertebral bodies and sternum. The orthogonal kV images were obtained with the on-board imaging (OBI) feature of the Trilogy System (Varian Medical Systems).
In addition, accuracy was assessed in a subset of patients who had internal fiducial markers (surgical clips) placed at the time of lumpectomy. Matching of orthogonal kV images to these markers served as the gold standard.
During each of nine fractions of radiation delivered over a week, patients were first aligned using the treatment room lasers. The investigators then obtained an AlignRT image and an OBI image, determining the projected offset or shift for each system. The patients were then shifted with the AlignRT system, and second AlignRT and OBI images were obtained.
Residual setup errors were ascertained by measuring the remaining offsets relative to a reference isocenter determined using the internal fiducial markers.
Analyses of shifts were based on 37 patients who received 331 fractions of radiation, Dr. Wahab said. Respective average shifts in the anterior-posterior, superior-inferior, and lateral axes were 4.6, 3.9, and 3.4 mm for AlignRT from laser alignment; 4.2, 3.5, and 4.2 mm for OBI from laser alignment; and 3.9, 4.5, and 4.0 mm for AlignRT vs OBI.
“This leads us to question which is the most appropriate surrogate for an internal target within a deformable structure such as the breast: Is it a surface map of skin or is it a deeper internal structure, perhaps a fixed structure like skeletal anatomy,” Dr. Wahab commented.
Analyses of accuracy were based on the subset of 18 patients with internal fiducial markers who received 160 fractions.
The respective average residual setup error values in the anterior-posterior, superior-inferior, and lateral axes were 1.9, 1.8, and 1.8 mm for AlignRT, with a summary value of 2 mm or less; 3.2, 4.2, and 4.7 mm for OBI, with a summary value of 5 mm or less; and 3.9, 4.6, and 4.3 mm for laser alignment, also with a summary value of 5 mm or less. When plotted in a histogram, the data showed a sharper peak and a narrower spread of values for AlignRT than for OBI (see Figure).
Finally, in analyses of reproducibility, the mean offsets between the first and second AlignRT images were 1.6 mm in the anterior-posterior axis, 1.3 mm in the superior-inferior axis, and 0.9 mm in the lateral axis. The respective mean offsets between the first and second OBI images were 2.6, 2.7, and 1.9 mm.
“So we can see that the AlignRT system is not only more accurate, but also more reproducible,” Dr. Wahab commented.
In conclusion, he stated, “kV imaging matching to skeletal anatomy was accurate only to within 5 mm, which in our study was no better than setup by standard alignment with lasers alone. We therefore recommend that if you are using orthogonal imaging, internal fiducial markers should be placed in your patients.”
He noted that surface mapping provided isocenter positioning accuracy to within 2 mm. “We found this to be satisfactory, and we believe that when using surface mapping, you do not need internal fiducial markers,” he said.
Dr. Taghian’s commentary
“There is no doubt that the big advantage of 3D conformal breast irradiation over MammoSite and implants is that it is a noninvasive procedure. But the big disadvantage is uncertainty of the setup and the error that might happen,” commented Alphonse Taghian, MD, PhD, associate professor of radiation oncology at Harvard and associate in radiation oncology at Massachusetts General Hospital.
In the study, residual setup error was less with AlignRT than with either lasers or OBI, although still no more than 5 mm with the latter methods, he noted.
The bottom line
“The bottom line from the study is that if you have an OBI system and fiducial markers, you will do fine by using the OBI. If you don’t have the fiducial markers, AlignRT is likely to be superior,” Dr. Taghian commented.
He added that “these types of studies should go hand in hand with studies investigating the 3D conformal technique when used in PBI.”