Radiation and Cancer Physics

SS 01 - Physics 1 -Best of Physics

14 - The Hippocampal NTCP Model Could Not be Validated Within the EORTC-22033 Low-Grade Glioma Trial

Sunday, October 21
2:35 PM - 2:45 PM
Location: Room 214 A/B

The Hippocampal NTCP Model Could Not be Validated Within the EORTC-22033 Low-Grade Glioma Trial
J. Jaspers1, A. Mendez Romero2, M. S. Hoogeman3, M. J. van den Bent4, R. Wiggenraad5, M. J. Taphoorn6, D. B. Eekers7, F. Lagerwaard8, A. Lucas9, B. G. Baumert10,11, and M. Klein12; 1Department of Radiation Oncology, Erasmus MC Cancer Institute, Rotterdam, Netherlands, 2Erasmus MC Daniel den Hoed Cancer Center, Rotterdam 3075 EA, Netherlands, 3Erasmus MC Cancer Institute, Rotterdam, Netherlands, 4Department of Neurology, Erasmus MC Cancer Institute, Rotterdam, Netherlands, 5Department of Radiation Oncology, Haaglanden Medical Center, The Hague, Netherlands, 6Department of Neurology, Haaglanden Medical Center, The Hague, Netherlands, 7Department of Radiation Oncology, MAASTRO Clinic, Maastricht, Netherlands, 8VU University Medical Center, Amsterdam, Netherlands, 9Department of Radiation Oncology, Institut Català d'Oncologia - L'Hospitalet, Barcelona, Spain, 10Department of Radiation Oncology, Paracelsus Clinic Osnabrück, Osnabrück, Germany, 11GROW - School for Oncology, Maastricht University Medical Centre, Maastricht, Netherlands, 12Department of Medical Psychology, VU University Medical Center, Amsterdam, Netherlands

Purpose/Objective(s): Comparative treatment planning using normal tissue complication probability (NTCP) evaluation has been proposed to select patients for proton therapy. To use NTCP models for selection, external validation is highly desirable. Here, we validated in a population of low-grade glioma (LGG) patients the NTCP model (Gondi et al.) that relates the dose to 40% of the bilateral hippocampus to neurocognitive impairment assessed at 18 months after treatment.

Materials/Methods: LGG patients treated within the radiotherapy-arm of the EORTC-22033 trial without evidence of clinical progression, having completed neurocognitive testing at baseline and at 18 months post-treatment, and having retrievable radiotherapy data, were included in this study. Patients were recruited in one Spanish and four Dutch hospitals. The clinical dose distributions were converted to EQD2Gy. To determine neurocognitive decline, differences in Rey Verbal Auditory Learning test delayed recall (AVLT-dr) performance pre and post-treatment were compared to test - retest data from healthy volunteers in the Maastricht Aging study. A score difference of -2.4 points was considered an event. NTCP values were calculated using the dose received by 40% of the bilateral hippocampus and compared to the observed neurocognitive outcome. Model performance was measured by the area under the receiver-operator curve (AUC) and the Brier score.

Results: A total of 29 patients met inclusion criteria. Mean EQD2Gy to 40% of the bilateral hippocampus was 39.8 Gy (34.3 – 44.4 Gy). The model predicted a risk of neurocognitive impairment exceeding 99% in 22 patients. At this point in time only in 5 of these patients a deterioration was observed. No significant difference was found between the pre-and post-treatment ALVT-dr score (mean score 9.28 vs 9.83, p = 0.48). The AUC was 0.57 and the Brier score was 0.67 demonstrating poor model performance.

Conclusion: The hippocampal NTCP model did not predict neurocognitive impairment at 18 months in LGG patients treated with radiotherapy within the EORTC 22033 trial. The model overestimates the risk of neurocognitive deterioration at 18 months in this patient group, while our observed rate of neurocognitive impairment is in line with published studies on LGG also using a short follow up. These findings do not support the use of this model to select LGG patients for proton beam therapy.
Age (years) Dose to 40% of bilateral hippocampus (EQD2Gy) NTCP prediction Cognitive event
48.7 3.21 0.07 no
48.0 7.30 0.17 no
36.3 10.04 0.27 no
69.4 18.45 0.67 yes
49.2 19.31 0.71 no
45.5 27.24 0.94 no
42.3 28.89 0.96 no
40.3 40.62 >0.99 no
32.9 44.30 >0.99 no
50.8 45.51 >0.99 no
37.1 46.27 >0.99 yes
35.6 46.52 >0.99 no
40.6 46.79 >0.99 no
41.6 47.08 >0.99 no
50.1 47.18 >0.99 yes
34.7 47.31 >0.99 no
48.5 47.42 >0.99 no
60.2 47.50 >0.99 yes
35.9 47.61 >0.99 no
36.3 47.87 >0.99 no
42.5 47.91 >0.99 no
29.5 47.91 >0.99 no
35.2 48.00 >0.99 no
66.5 48.13 >0.99 no
44.3 48.20 >0.99 yes
34.0 48.43 >0.99 no
50.6 48.73 >0.99 yes
32.0 48.95 >0.99 no
27.8 50.74 >0.99 no

Author Disclosure: J. Jaspers: None. A. Mendez Romero: None. M.S. Hoogeman: Board member; Accuray Exchange in Radiation Oncology. M.J. van den Bent: None. R. Wiggenraad: None. D.B. Eekers: None. A. Lucas: None. B.G. Baumert: None.

Jaap Jaspers, MD

Disclosure:
No relationships to disclose.

Presentation(s):

Send Email for Jaap Jaspers


Assets

14 - The Hippocampal NTCP Model Could Not be Validated Within the EORTC-22033 Low-Grade Glioma Trial

Slides

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for The Hippocampal NTCP Model Could Not be Validated Within the EORTC-22033 Low-Grade Glioma Trial