Glioblastoma and radiotherapy: A multicenter AI study for Survival Predictions from MRI (GRASP study).

Chelliah, A.; Wood, DA.; Canas, LS.; Shuaib, H.; Currie, S.; Fatania, K.; Frood, R.; Rowland-Hill, C.; Thust, S.; Wastling, SJ.; Tenant, S.; McBain, C.; Foweraker, K.; Williams, M.; Wang, Q.; Roman, A.; Dragos, C.; MacDonald, M.; Lau, YH.; Linares, CA.; Bassiouny, A.; Luis, A.; Young, T.; Brock, J.; Chandy, E.; Beaumont, E.; Lam, TC.; Welsh, L.; Lewis, J.; Mathew, R.; Kerfoot, E.; Brown, R.; Beasley, D.; Glendenning, J.; Brazil, L.; Swampillai, A.; Ashkan, K.; Ourselin, S.; Modat, M.; Booth, TC.

Glioblastoma and radiotherapy: A multicenter AI study for Survival Predictions from MRI (GRASP study).

All Authors

Chelliah, A.

Wood, DA.

Canas, LS.

Shuaib, H.

Currie, S.

Fatania, K.

Frood, R.

Rowland-Hill, C.

Thust, S.

Wastling, SJ.

Show 10 more

LTHT Author

Currie, Stuart
Fatania, Kavi
Frood, Russell
Mathew, Ryan

LTHT Department

Radiology
Neurosurgery

Publication Date

2024

Item Type

Journal Article
Multicenter Study
Research Support, Non-U.S. Gov't

Abstract

BACKGROUND: The aim was to predict survival of glioblastoma at 8 months after radiotherapy (a period allowing for completing a typical course of adjuvant temozolomide), by applying deep learning to the first brain MRI after radiotherapy completion. METHODS: Retrospective and prospective data were collected from 206 consecutive glioblastoma, isocitrate dehydrogenase -wildtype patients diagnosed between March 2014 and February 2022 across 11 UK centers. Models were trained on 158 retrospective patients from 3 centers. Holdout test sets were retrospective (n = 19; internal validation), and prospective (n = 29; external validation from 8 distinct centers). Neural network branches for T2-weighted and contrast-enhanced T1-weighted inputs were concatenated to predict survival. A nonimaging branch (demographics/MGMT/treatment data) was also combined with the imaging model. We investigated the influence of individual MR sequences; nonimaging features; and weighted dense blocks pretrained for abnormality detection. RESULTS: The imaging model outperformed the nonimaging model in all test sets (area under the receiver-operating characteristic curve, AUC P = .038) and performed similarly to a combined imaging/nonimaging model (P > .05). Imaging, nonimaging, and combined models applied to amalgamated test sets gave AUCs of 0.93, 0.79, and 0.91. Initializing the imaging model with pretrained weights from 10 000s of brain MRIs improved performance considerably (amalgamated test sets without pretraining 0.64; P = .003). CONCLUSIONS: A deep learning model using MRI images after radiotherapy reliably and accurately determined survival of glioblastoma. The model serves as a prognostic biomarker identifying patients who will not survive beyond a typical course of adjuvant temozolomide, thereby stratifying patients into those who might require early second-line or clinical trial treatment.