Assessment of organs-at-risk contouring practices in radiosurgery institutions around the world - The first initiative of the OAR Standardization Working Group

Helena Sandström; Caroline Chung; Hidefumi Jokura; Michael Torrens; David Jaffray; Iuliana Toma-Dasu

doi:10.1016/j.radonc.2016.10.014

Assessment of organs-at-risk contouring practices in radiosurgery institutions around the world - The first initiative of the OAR Standardization Working Group

Radiother Oncol. 2016 Nov;121(2):180-186. doi: 10.1016/j.radonc.2016.10.014. Epub 2016 Nov 2.

Authors

Helena Sandström¹, Caroline Chung², Hidefumi Jokura³, Michael Torrens⁴, David Jaffray⁵, Iuliana Toma-Dasu⁶

Affiliations

¹ Medical Radiation Physics, Stockholm University and Karolinska Institutet, Sweden. Electronic address: helena.sandstrom@ki.se.
² Department of Radiation Oncology, University of Toronto and University Health Network-Princess Margaret Cancer Centre, Canada.
³ Suzuki Memorial Gamma House, Furukawa Seiryo Hospital, Osaki, Japan.
⁴ Gamma Knife Department, Hygeia Hospital, Athens, Greece.
⁵ Department of Radiation Oncology, Medical Biophysics, IBBME, University of Toronto and University Health Network-Princess Margaret Cancer Centre/TECHNA, Canada.
⁶ Medical Radiation Physics, Stockholm University and Karolinska Institutet, Sweden.

PMID: 27816408
DOI: 10.1016/j.radonc.2016.10.014

Abstract

Background and purpose: This study was an initiative of the Organs-at-Risk Standardization Working Group for evaluating the current degree of variability in the clinical practice of contouring organs-at-risk (OAR) for radiosurgery planning.

Materials and methods: Imaging datasets for typical lesions (cavernous sinus meningioma, vestibular schwannoma, pituitary adenoma) treated with Leksell Gamma Knife Perfexion were circulated to 12 centers. Observers were asked to contour the target and OARs as per their standard clinical practice. The analyzed parameters were the intersection (AV₁₀₀), union volumes (AV_100/N) and the 50% agreement volume (AV₅₀). The ratio of AV₁₀₀ and AV_100/N (the Agreement Volume Index, AVI) was used as a measure of agreement level together with a generalized conformity index (CI_gen) and a pairwise averaged conformity index (CI_pairs). The maximum doses were also determined.

Results: Results showed a wide variability in terminology, choice of structures contoured and in the size and shape of the contoured structures. The highest variability was observed for the left and right optic tract for cavernous sinus meningioma where the AV₁₀₀ was zero. The highest consistency was observed for the right optic nerve in the cavernous sinus case followed by the cochlea for the vestibular schwannoma case for which the AVI was still only 0.13 and 0.054, respectively. Corresponding results for the CI_gen and CI_pairs also showed the highest variability for the right optic tract and the highest consistency in contours for the right optic nerve, both in the cavernous sinus meningioma case.

Conclusion: The results quantify the large variability in OAR contouring in clinical practice across Gamma Knife radiosurgery centers with respect to the choice of OARs to be contoured, nomenclature and size and shape of OARs. This motivates future effort to standardize practices to enable more effective collaboration.

Keywords: Chiasm; Cochlea; Inter-observer variability; Optic nerve; Stereotactic radiosurgery.

MeSH terms

Cavernous Sinus / pathology
Humans
Magnetic Resonance Imaging
Meningeal Neoplasms / surgery
Meningioma / radiotherapy
Neuroma, Acoustic / radiotherapy
Optic Nerve / pathology
Organs at Risk* / pathology
Pituitary Neoplasms / radiotherapy
Radiosurgery* / methods