Interfraction and intrafraction performance of the Gamma Knife Extend system for patient positioning and immobilization.

OBJECT: The Extend system for the Gamma Knife Perfexion makes possible multifractional Gamma Knife treatments. The Extend system consists of a vacuum-monitored immobilization frame and a positioning measurement system used to determine the location of the patient's head within the frame at the time of simulation imaging and before each treatment fraction. The measurement system consists of a repositioning check tool (RCT), which attaches to the Extend frame, and associated digital measuring gauges. The purpose of this study is to evaluate the performance of the Extend system for patient repositioning before each treatment session (fraction) and patient immobilization between (interfraction) and during (intrafraction) each session in the first 10 patients (36 fractional treatments) treated at the University of Virginia.
METHODS: The RCT was used to acquire a set of reference measurements for each patient position at the time of CT simulation. Repositioning measurements were acquired before each fraction, and the patient position was adjusted until the residual radial difference from the reference position measurements was less than 1 mm. After treatment, patient position measurements were acquired, and the difference between those measurements and the ones obtained for patient position before the fraction was calculated as a measure of immobilization capability. Analysis of patient setup and immobilization performance included calculation of the group mean, standard deviation (SD), and distribution of systematic (components affecting all fractions) and random (per fraction) uncertainty components.
RESULTS: Across all patients and fractions, the mean radial setup difference from the reference measurements was 0.64 mm, with an SD of 0.24 mm. The distribution of systematic uncertainty (Σ) was 0.17 mm, and the distribution of random uncertainty (σ) was 0.16 mm. The root mean square (RMS) differences for each plate of the RCT were as follows: right = 0.35 mm; left = 0.41 mm; superior = 0.28 mm; and anterior = 0.20 mm. The mean intrafractional positional difference across all treatments was 0.47 mm, with an SD of 0.30 mm. The distribution of systematic uncertainty was 0.18 mm, and the distribution of random uncertainty was 0.22 mm. The RMS differences for each plate of the RCT were 0.24 mm for the right plate, 0.22 mm for the left plate, 0.24 mm for the superior plate, and 0.34 mm for the anterior plate. Data from 1 fraction were excluded from the analysis because the vacuum-monitoring interlock detected patient motion, which in turn required repositioning in the middle of the fraction.
CONCLUSIONS: The Extend system can be used to reposition and immobilize patients in a radiosurgical setting. However, care should be taken to acquire measurements that can implicitly account for rotations of the patient's head. Further work is required to determine the sensitivity of the vacuum interlock to detect patient motion.