In silico assessment of the dosimetric quality of a novel, automated radiation treatment planning strategy for linac-based radiosurgery of multiple brain metastases and a comparison with robotic methods

Abstract

Background: To appraise the dosimetric features and the quality of the treatment plan for radiosurgery of multiple brain metastases optimized with a novel automated engine and to compare with plans optimized for robotic-based delivery.

Methods: A set of 15 patients with multiple brain metastases was selected for this in silico study. The technique under investigation is the recently introduced HyperArc. For all patients, three treatment plans were computed and compared: i: a HyperArc; ii: a standard VMAT; iii) a CyberKnife. Dosimetric features were computed for the clinical target volumes as well as for the healthy brain tissue and the organs at risk.

Results: The data showed that the best dose homogeneity was achieved with the VMAT technique. HyperArc allowed to minimize the volume of brain receiving 4Gy (as well as for the mean dose and the volume receiving 12Gy, although not statistically significant). The smallest dose on 1 cm³ volume for all organs at risk is for CK techniques, and the biggest for VMAT (p < 0.05). The Radiation Planning Index coefficient indicates that, there are no significant differences among the techniques investigated, suggesting an equivalence among these.

Conclusion: At treatment planning level, the study demonstrates that the use of HyperArc technique can significantly improve the sparing of the healthy brain while maintaining a full coverage of the target volumes.

Keywords: CyberKnife; HyperArc; Multiple brain metastases; Radiosurgery; VMAT.

Fig. 1

A screen shot of the user’s interface for the definition of the HyperArc treatment plan

Fig. 2

Conceptual visualization of the complexity of the manual arrangement of the collimator angle for the standard VMAT planning. On the contrary, the new HyperArc engine, includes an automatic optimization of the collimator angle

Fig. 3

Axial views of the dose distribution for an example case with three lesions for the three different techniques, arranged in columns: HyperArc, VMAT, CyberKnife. The colorwash is set to 13.5 Gy (90% of the prescription) to 17.0 Gy – upper row, 10.0 Gy to 17.0 Gy – middle row, 4.0 Gy to 17.0 Gy – lower row. It is qualitatively noticeable the different management of the dose bridging and of the gradients between the different approaches

Fig. 4

The cumulative dose volume histograms of the example case with three lesions for the three techniques under investigation. In red the total target volume, in green the healthy brain

Fig. 5

Summary of the analysis of the PTV data. a near-to-minimum dose; b homogeneity. The central point represents the mean, the solid box the standard error and the bars the standard deviation

Fig. 6

Summary of the analysis of the data for the organs at risk

Fig. 7

Summary of the analysis of the data for the healthy brain. In the first panel the mean dose is reported while in the second and in the third panels the V_12Gy and the V_4Gy data

Fig. 8

Summary of the analysis of the RPI. In the first panel when all the structures per each plan are included and in the second panel when only the targets and the healthy brain are considered