PV QA 3 - Poster Viewing Q&A 3
TU_21_3276 - Optimalization of Choice of the SUV of the PET(68)Ga-DOTATATE Imaging to Achieve Highest Precision for Contouring the GTV of Meningiomas in Accuracy Stereotactic Hypofractionated Radiosurgery
Tuesday, October 23
1:00 PM - 2:30 PM
Location: Innovation Hub, Exhibit Hall 3
Optimalization of Choice of the SUV of the PET (68)Ga-DOTATATE Imaging to Achieve Highest Precision for Contouring the GTV of Meningiomas in Accuracy Stereotactic Hypofractionated Radiosurgery
B. Maciejewski1, M. Stapor-Fudzinska2, and L. Hawrylewicz3; 1Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology Gliwice Branch, Gliwice, Poland, 2Department of Radiotherapy Planning, Maria Skłodowska-Curie Memorial Cancer Center and Institute of Oncology, Gliwice Branch, Gliwice, Poland, 3Institute of Oncology Gliwice, Poland, Gliwice, Poland
Purpose/Objective(s): In the previous study we recommended CT/PET (68)Ga-DOTATATE as a primary, sensitive and precise imaging to contour GTV of a large (> 10 ccm) meningiomas (Mngs) with irregular margins (~ 40%). The Mngs contours on CT/PET (68)Ga scans strongly depend on a SUV value. Aim of the present study is to estimate optimal cut-off range of the SUV providing the highest possible precision and conformity of the meningioma GTV contouring.
Materials/Methods: Seventy Mngs (55 pts) were treated with Robotic Stereotactic Hypofractionated Radiosurgery in a single institution. In the present study two independent medical physicists contoured Mng GTVs on serial CT/PET (68)Ga-DOTATATE scans using four different SUV ranges, i. e. 0-100%, 25-100%, 50-100% and 75-100% for each GTV. The PET voxel size was 0.054 ccm and serial scans were always separated by 3 mm. The GTV contoured by one observer was covered to the another. Interobserver conformity (CI), sharpness of the GTV contours and Volume Overlap Ratio (VOR) were evaluated for each of the four SUV ranges to estimate optimal cut-off range of the SUV.
Results: For small Mngs GTV (≤ 5 ccm) interobserver conformity of the GTV contours was high (CI = 0.85) but much lower (CI = 0.59) for GTV larger than 10 ccm. In the group of 13 Mngs pronounced variations in the GTV contouring were noted mainly due to irregular and misty GTV margins. Analysis of relationship between SUV range and Mngs GTV contours showed that with increasing the SUV from 0-100% to 75-100% the GTV contour reduces towards the center of the GTV together with intensive lightened zones as well, what leads to the increasing uncertainty of the real GTV contour. For SUV of 0-100% and 25-100% sharp GTV margins were noted in 94% of the Mngs and this rate decreased to only 5% when the SUV value raised to 50-100% and 75-100%. Analysis of the VOR index (Volume Overlap Ratio is a ratio of GTVs contoured using SUV 25-, 50-, 75-100% related to the reference GTV contours for SUV 0-100%) showed that the respective GTVs contours overlap smaller and smaller part of the reference GTV and about 20-70% of the reference volumes were left out when the GTVs were contoured using the SUV of 50-100% and 75-100%. Analysis of the interobserver conformity index of the CI > 0.7 characterized two-times more Mngs GTVs contours using SUV of 25-100% than in the case the SUV of 0-100%. Low CI of < 0.5 was found in about 30% of GTVs contours using SUV 0-100% and in zero % of GTV for the SUV of 25-100%. Therefore, the SUV range of 25-100% is recommended as optimal cut-off value.
Conclusion: The SUV-PET (68)Ga-DOTATATE of 25-100% is recommended as optimal for contouring meningioma’s GTV margins because it ensures the highest possible GTV contour’s conformity. Visible GTV contours on PET (68)Ga with the SUV of 50-100% and 75-100% are significantly reduced and a large part of real GTV was left beyond its margins and therefore it is not advised to use the SUV 50-100% and 75-100% in practice.
Author Disclosure: B. Maciejewski: None. M. Stapor-Fudzinska: None.