ESTRO 2024 - Abstract Book

S842

Clinical - CNS

ESTRO 2024

common primary histology (77%) followed by breast cancer (17%). All patients were treated using Gammaknife based SRS to 20-24 Gy. Eighteen (38%) were treated for a single metastasis, 25 (52%) for 2-10 metastases, and 5 (10%) for >10 metastases. At the 3- month assessment, ND (≥2 SE) in any 1 test was observed in 40 (85%) and in ≥2 tests in 29 (62%): verbal recall in 40%, verbal revision in 32%, verbal recognition in 27%, verbal fluency in 36%, numbers ordering in 6%, numbers and letters ordering in 25%, symbols matching in 30%. Using a more stringent threshold of ≥3 SE in any 1 test, ND was observed in 37 (79%) and in ≥2 tests in 26 (55%): verbal recall in 36%, verbal revision in 32%, verbal recognition in 27%, verbal fluency in 36%, numbers ordering in 6%, numbers and alphabet ordering in 23%, symbols matching in 23%. At the 6- month assessment (n=20 patients), ND (≥ 2 SE) in any 1 test was observed in 19 (95%) and in ≥2 tests in 14 (70%); using ≥3 SE threshold, ND in any 1 test was observed in 16 (80%) and in ≥2 tests in 12 (60%). Across all patients, dosimetric analysis revealed low hippocampal doses across all metrics: median of mean U/L hippocampus dose was 0.14 Gy (R: 0-2.72 Gy), median of mean B/L hippocampus dose was 0.17 Gy (R: 0-2.38 Gy). The median U/L D100 was 0.03 Gy (R: 0-1.24 Gy), median B/L D100 was 0.03 Gy (R: 0-1.21), median U/L D40 was 0.14 Gy (R: 0-2.86), median B/L D40 was 0.17 Gy (R: 0-2.53), median U/L D0.03cc was 0.4 Gy (R: 0-5.57 Gy), and median B/L D0.03cc was 0.51 Gy (R: 0-5.55 Gy). Logistic regression for hippocampal dosimetric predictors of ND in verbal recognition at 3 months revealed that U/L D100 and D40 were significantly associated (p=0.04). Analysis based on the number of metastases treated revealed significant differences in U/L D100 (median: 0.01, 0.05, 0.22 for 1 metastasis, 2-10 metastases, and > 10 metastases, respectively [p<0.01]) and D40 between the groups (median: 0.04, 0.18, 0.78 for 1 metastasis, 2-10 metastases, and > 10 metastases, respectively [p<0.01]).

Conclusion:

In this unique prospective app-based assessment, relatively high rates of neurocognitive decline were detected in patients treated with SRS for brain metastasis despite very low hippocampal dosimetric parameters. Unilateral D100 and D40 were associated with neurocognitive decline at 3 months using this assessment, recapitulating results from traditional pen and paper-based assessments. Differences in these parameters were also observed in patients treated with multiple brain metastases compared to those treated for a single brain metastasis.

Keywords: Neurocognitive outcomes, SRS, Brain metastasis

1474

Digital Poster

HYPOFRACTIONATED

VS

CONVENTIONAL

FRACTIONATED

RADIOTHERAPY

FOR

GLIOBLASTOMA:INSTITUTIONAL EXPERIENCE

FIORELLA CRISTINA DI GUGLIELMO 1 , FABIANA GREGUCCI 1 , MARIA PAOLA CILIBERTI 1 , VALERIO DAVì 1 , ILARIA BONAPARTE 2 , ALESSIA SURGO 3 , MORENA CALIANDRO 3 , ROBERTA CARBONARA 3 , ALBA FIORENTINO 3,4 1 Miulli General Regional Hospital, Department of Radiation Oncology, Acquaviva delle Fonti, Italy. 2 Miulli General Regional Hospital, Department of Radiation Oncology, ACQUAIVA DELLE FONTI, Italy. 3 Miulli General Regional