Brain Diffusion Tensor Imaging to Predict Immunotherapy Response in Metastatic Melanoma (britemet): a Single-centre Biomarker Pilot Study
Year: 2025
Type: conference-abstract
Abstract: AIMS Around half of patients with brain metastases respond to immune checkpoint inhibition and these treatments have significant side effects, highlighting the critical need for treatment response biomarkers. Diffusion Tensor Imaging (DTI) represents a rapidly obtained MRI sequence that has been shown to discriminate high from low peritumoural immune infiltration and could therefore help personalise treatment approaches. METHODS This prospective single-centre biomarker study (CRUK funded) recruited adult patients with intracranial melanoma metastases eligible for immunotherapy. All patients received standard ipilimumab/nivolumab com- bination immunotherapy. We assessed pre-treatment DTI and 6-week follow-up scans using RANO and RE- CIST criteria. Peritumoral fractional anisotropy (FA) values from the DTI scans were measured blind to clin- ical outcomes using both manual (two raters; one being a neuroradiologist with 10+ years experience) and semi-automated techniques. We performed receiver operating characteristic (ROC) curve analysis to assess the discriminative performance of FA measurements in predicting both intracranial and extracranial response, calculating area under the ROC curve (AUC) values for both measurement techniques. RESULTS Eight patients with 18 brain metastases were included. Median overall survival was 13.1 months (95% CI: 1.6-24.6), with median intracranial progression-free survival of 12.9 months (95% CI: 7.2-18.6) and median ex- tracranial progression-free survival of 14.3 months (95% CI: 11.7-16.9). Both manual and semi-automated FA measurements demonstrated excellent predictive value for brain metastasis response (AUC=0.85 and 0.825, re- spectively), while showing poor discrimination for extracranial response (AUC=0.333 and 0.556, respectively). There was discordance between automated and manual FA readings but excellent inter-rater agreement for manual recording of FA. CONCLUSION This study demonstrates that DTI-derived FA measurements show promise as a predictive biomarker for im- munotherapy response in brain metastases, providing a potentially useful tool for optimizing patient selection. Further refinement of the semi-automated technique for measurement (e.g. via a plugin for PACS) and valida- tion in a larger cohort with diverse primary tumour types is needed
Source: Neuro-Oncology
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4413927795