Overall Accuracy of Single-Isocenter Multiple Metastases SRS Treatments over Time: Comparison of Two Commercially Available Methods

Abstract

<h3>Purpose/Objective(s)</h3> Single isocenter stereotactic radiosurgery (SI-SRS) is an efficient approach for dose delivery to multiple brain metastases since the treatment duration is considerably reduced, compared to the standard method of the isocenter placement at each tumor center. However, the complexity of this technique results in increased sensitivity to geometric and dosimetric uncertainties during plan delivery, thus patient-specific quality assurance procedures of high accuracy are mandatory. This study presents the investigation into the consistency of this technique, using the a 4D dose verification system and 3D printed head phantoms. <h3>Materials/Methods</h3> A treatment plan with two brain metastases was delivered four times in the period of one month on a Linac, to evaluate the reproducibility of the SI-SRS technique. The targets had a concave (2.4 cc) and a spherical (1.5 cc) shape, placed at a distance of 5 cm. Dose prescription was set to 8 Gy for both targets and a treatment plan of a single fraction using four Hyperarc beams, one full coplanar arc and three half noncoplanar arcs, arranged with a single isocenter automatically located on the center of the distance between the two targets was created. Before plan delivery, Winston-Lutz (WL) tests were performed to determine the diameter of the sphere containing the radiation isocenter for various positions of the collimator, gantry and couch. The dose distribution was recalculated on 4D geometry and measured with the 1000 SRS ion chamber array. The head phantom accommodated a film insert and the irradiation procedure consisted of positioning, imaging and plan delivery as if it was the real patient. Gamma analysis was performed for 3%/1mm acceptance criteria using a dose threshold of 1 Gy for films and 10% of maximum dose for 4D. <h3>Results</h3> The Root Mean Square (RMS) variation of the isocenter was 0.71±0.15 mm for gantry, 0.66±0.14 mm for collimator and 0.87±0.22 mm for couch, validating that the isocenter diameter was within the tolerance of 1 mm before each measurement. Gamma analysis revealed a mean passing rate (GPR) of 93.2%±0.07% for 4D measurements, compared to 91.2%±1.8% for the film ones. The reduced GPR in the case of the head phantom is expected, since it is an End-to-End QA procedure, incorporating not only the plan delivery uncertainties, but also the ones related to image registration and couch rotation. <h3>Conclusion</h3> Both QA methods showed a good agreement with each other and the calculated dose distribution having GPRs above 90% for the 3%/1mm criteria, rendering them suitable for verification of SRS plans' dosimetric and spatial accuracy. The standard deviation of the GPRs revealed that the plan delivery was consistent between the measurements, demonstrating high reproducibility and accuracy of the SI-SRS technique.